File size: 68,154 Bytes
96b4732 a534622 96b4732 5347d74 a731957 47ed41d a731957 ae2d783 96b4732 5347d74 a731957 ae2d783 a731957 ae2d783 a731957 ae2d783 a731957 ae2d783 a731957 ae2d783 150d26a ae2d783 a731957 ae2d783 a731957 ae2d783 a731957 5347d74 96b4732 5347d74 96b4732 5347d74 96b4732 5347d74 96b4732 a731957 96b4732 5347d74 a731957 5347d74 a731957 5347d74 a731957 5347d74 a731957 96b4732 5347d74 96b4732 a534622 133034a a534622 ae2d783 a534622 96b4732 ddcd358 96b4732 b621ea4 a731957 b621ea4 96b4732 b621ea4 96b4732 150d26a b621ea4 150d26a 96b4732 a731957 96b4732 cea43a3 150d26a a534622 96b4732 cea43a3 96b4732 a731957 96b4732 150d26a 96b4732 150d26a 96b4732 cea43a3 96b4732 a534622 b621ea4 a534622 5383ed1 96b4732 52e62db 96b4732 cea43a3 96b4732 cea43a3 96b4732 cea43a3 96b4732 829271f cea43a3 96b4732 133034a 96b4732 829271f 96b4732 cea43a3 96b4732 133034a 96b4732 b621ea4 96b4732 a731957 96b4732 a731957 96b4732 5347d74 96b4732 a731957 96b4732 5a42dc4 96b4732 5a42dc4 96b4732 5a42dc4 ddcd358 5a42dc4 96b4732 ec47763 96b4732 ec47763 96b4732 ec47763 96b4732 ec47763 96b4732 5347d74 cea43a3 5347d74 cea43a3 96b4732 5347d74 cea43a3 d82e9a1 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 |
import streamlit as st
import streamlit.components.v1 as components
import os
import json
import random
import base64
import glob
import math
import openai
import pytz
import re
import requests
import textract
import time
import zipfile
import dotenv
from gradio_client import Client
from audio_recorder_streamlit import audio_recorder
from bs4 import BeautifulSoup
from collections import deque
from datetime import datetime
from dotenv import load_dotenv
from huggingface_hub import InferenceClient
from io import BytesIO
from openai import ChatCompletion
from PyPDF2 import PdfReader
from templates import bot_template, css, user_template
from xml.etree import ElementTree as ET
from PIL import Image
from urllib.parse import quote # Ensure this import is included
# 1. Configuration
Site_Name = 'Scholarly-Article-Document-Search-With-Memory'
title="🚀🌌ArXiv Article Document Search Memory"
helpURL='https://huggingface.co/awacke1'
bugURL='https://huggingface.co/spaces/awacke1'
icons='🔍🚀🌌📖'
st.set_page_config(
page_title=title,
page_icon=icons,
layout="wide",
initial_sidebar_state="expanded",
menu_items={
'Get Help': helpURL,
'Report a bug': bugURL,
'About': title
}
)
def load_file(file_name):
with open(file_name, "r", encoding='utf-8') as file:
#with open(file_name, "r") as file:
content = file.read()
return content
# HTML5 based Speech Synthesis (Text to Speech in Browser)
@st.cache_resource
def SpeechSynthesis(result):
documentHTML5='''
<!DOCTYPE html>
<html>
<head>
<title>Read It Aloud</title>
<script type="text/javascript">
function readAloud() {
const text = document.getElementById("textArea").value;
const speech = new SpeechSynthesisUtterance(text);
window.speechSynthesis.speak(speech);
}
</script>
</head>
<body>
<h1>🔊 Read It Aloud</h1>
<textarea id="textArea" rows="10" cols="80">
'''
documentHTML5 = documentHTML5 + result
documentHTML5 = documentHTML5 + '''
</textarea>
<br>
<button onclick="readAloud()">🔊 Read Aloud</button>
</body>
</html>
'''
components.html(documentHTML5, width=1280, height=300)
def parse_to_markdown(text):
return text
import re
def extract_urls(text):
try:
# Regular expression patterns to find the required fields
date_pattern = re.compile(r'### (\d{2} \w{3} \d{4})')
abs_link_pattern = re.compile(r'\[(.*?)\]\((https://arxiv\.org/abs/\d+\.\d+)\)')
pdf_link_pattern = re.compile(r'\[⬇️\]\((https://arxiv\.org/pdf/\d+\.\d+)\)')
title_pattern = re.compile(r'### \d{2} \w{3} \d{4} \| \[(.*?)\]')
# Find all occurrences of the required fields using the regular expression patterns
date_matches = date_pattern.findall(text)
abs_link_matches = abs_link_pattern.findall(text)
pdf_link_matches = pdf_link_pattern.findall(text)
title_matches = title_pattern.findall(text)
# Generate markdown string with the extracted fields
markdown_text = ""
for i in range(len(date_matches)):
date = date_matches[i]
title = title_matches[i]
abs_link = abs_link_matches[i][1]
pdf_link = pdf_link_matches[i]
markdown_text += f"**Date:** {date}\n\n"
markdown_text += f"**Title:** {title}\n\n"
markdown_text += f"**Abstract Link:** [{abs_link}]({abs_link})\n\n"
markdown_text += f"**PDF Link:** [{pdf_link}]({pdf_link})\n\n"
markdown_text += "---\n\n"
return markdown_text
except:
st.write('.')
return ''
def download_pdfs(urls):
local_files = []
for url in urls:
if url.endswith('.pdf'):
local_filename = url.split('/')[-1]
response = requests.get(url)
with open(local_filename, 'wb') as f:
f.write(response.content)
local_files.append(local_filename)
return local_files
def generate_html(local_files):
html = "<ul>"
for file in local_files:
link = f'<li><a href="{file}">{file}</a></li>'
html += link
html += "</ul>"
return html
#@st.cache_resource
def search_arxiv(query):
start_time = time.strftime("%Y-%m-%d %H:%M:%S")
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
search_query = query
search_source = "Arxiv Search - Latest - (EXPERIMENTAL)"
llm_model = "mistralai/Mixtral-8x7B-Instruct-v0.1"
#st.markdown('### 🔎 ' + query)
# Search 1 - Retrieve the Papers
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
response1 = client.predict(
query,
20,
"Semantic Search - up to 10 Mar 2024",
"mistralai/Mixtral-8x7B-Instruct-v0.1",
api_name="/update_with_rag_md"
)
lastpart = ''
totalparts = ''
Question = '### 🔎 ' + query + '\r\n' # Format for markdown display with links
References = response1[0]
References2 = response1[1]
#st.markdown(results)
# URLs from the response
ReferenceLinks = extract_urls(References)
#st.markdown(urls)
#results = results + urls
RunSecondQuery = True
if RunSecondQuery:
# Search 2 - Retrieve the Summary with Papers Context and Original Query
response2 = client.predict(
query,
"mistralai/Mixtral-8x7B-Instruct-v0.1",
True,
api_name="/ask_llm"
)
#st.markdown(response2)
if len(response2) > 10:
Answer = response2
SpeechSynthesis(Answer)
# Restructure results to follow format of Question, Answer, References, ReferenceLinks
results = Question + '\r\n' + Answer + '\r\n' + References + '\r\n' + ReferenceLinks
st.markdown(results)
st.write('🔍Run of Multi-Agent System Paper Summary Spec is Complete')
end_time = time.strftime("%Y-%m-%d %H:%M:%S")
# Output
start_timestamp = time.mktime(time.strptime(start_time, "%Y-%m-%d %H:%M:%S"))
end_timestamp = time.mktime(time.strptime(end_time, "%Y-%m-%d %H:%M:%S"))
elapsed_seconds = end_timestamp - start_timestamp
st.write(f"Start time: {start_time}")
st.write(f"Finish time: {end_time}")
st.write(f"Elapsed time: {elapsed_seconds:.2f} seconds")
#SpeechSynthesis(results)
filename = generate_filename(query, "md")
create_file(filename, query, results, should_save)
return results
def download_pdfs_and_generate_html(urls):
pdf_links = []
for url in urls:
if url.endswith('.pdf'):
pdf_filename = os.path.basename(url)
download_pdf(url, pdf_filename)
pdf_links.append(pdf_filename)
local_links_html = '<ul>'
for link in pdf_links:
local_links_html += f'<li><a href="{link}">{link}</a></li>'
local_links_html += '</ul>'
return local_links_html
def download_pdf(url, filename):
response = requests.get(url)
with open(filename, 'wb') as file:
file.write(response.content)
# Show ArXiv Scholary Articles! ----------------*************----▶️ Semantic and Episodic Memory System
def search_arxiv_old(query):
start_time = time.strftime("%Y-%m-%d %H:%M:%S")
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
search_query = query
search_source = "Arxiv Search - Latest - (EXPERIMENTAL)" # "Semantic Search - up to 10 Mar 2024"
llm_model = "mistralai/Mixtral-8x7B-Instruct-v0.1"
st.markdown('### 🔎 ' + query)
# Search 1 - Retrieve the Papers
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
response1 = client.predict(
query,
20,
"Semantic Search - up to 10 Mar 2024", # Literal['Semantic Search - up to 10 Mar 2024', 'Arxiv Search - Latest - (EXPERIMENTAL)'] in 'Search Source' Dropdown component
"mistralai/Mixtral-8x7B-Instruct-v0.1", # Literal['mistralai/Mixtral-8x7B-Instruct-v0.1', 'mistralai/Mistral-7B-Instruct-v0.2', 'google/gemma-7b-it', 'None'] in 'LLM Model' Dropdown component
api_name="/update_with_rag_md"
)
lastpart=''
totalparts=''
results = response1[0] # Format for markdown display with links
results2 = response1[1] # format for subquery without links
st.markdown(results)
RunSecondQuery = False
if RunSecondQuery:
# Search 2 - Retieve the Summary with Papers Context and Original Query
response2 = client.predict(
query, # str in 'parameter_13' Textbox component
"mistralai/Mixtral-8x7B-Instruct-v0.1",
#"mistralai/Mistral-7B-Instruct-v0.2",
#"google/gemma-7b-it",
True, # bool in 'Stream output' Checkbox component
api_name="/ask_llm"
)
st.markdown(response2)
results = results + response2
st.write('🔍Run of Multi-Agent System Paper Summary Spec is Complete')
end_time = time.strftime("%Y-%m-%d %H:%M:%S")
start_timestamp = time.mktime(time.strptime(start_time, "%Y-%m-%d %H:%M:%S"))
end_timestamp = time.mktime(time.strptime(end_time, "%Y-%m-%d %H:%M:%S"))
elapsed_seconds = end_timestamp - start_timestamp
st.write(f"Start time: {start_time}")
st.write(f"Finish time: {end_time}")
st.write(f"Elapsed time: {elapsed_seconds:.2f} seconds")
SpeechSynthesis(results) # Search History Reader / Writer IO Memory - Audio at Same time as Reading.
filename=generate_filename(query, "md")
create_file(filename, query, results, should_save)
return results
# Prompts for App, for App Product, and App Product Code
PromptPrefix = 'Create a specification with streamlit functions creating markdown outlines and tables rich with appropriate emojis for methodical step by step rules defining the concepts at play. Use story structure architect rules to plan, structure and write three dramatic situations to include in the rules and how to play by matching the theme for topic of '
PromptPrefix2 = 'Create a streamlit python user app with full code listing to create a UI implementing the using streamlit, gradio, huggingface to create user interface elements like emoji buttons, sliders, drop downs, and data interfaces like dataframes to show tables, session_statematching this ruleset and thematic story plot line: '
PromptPrefix3 = 'Create a HTML5 aframe and javascript app using appropriate libraries to create a word game simulation with advanced libraries like aframe to render 3d scenes creating moving entities that stay within a bounding box but show text and animation in 3d for inventory, components and story entities. Show full code listing. Add a list of new random entities say 3 of a few different types to any list appropriately and use emojis to make things easier and fun to read. Use appropriate emojis in labels. Create the UI to implement storytelling in the style of a dungeon master, with features using three emoji appropriate text plot twists and recurring interesting funny fascinating and complex almost poetic named characters with genius traits and file IO, randomness, ten point choice lists, math distribution tradeoffs, witty humorous dilemnas with emoji , rewards, variables, reusable functions with parameters, and data driven app with python libraries and streamlit components for Javascript and HTML5. Use appropriate emojis for labels to summarize and list parts, function, conditions for topic:'
roleplaying_glossary = {
"🤖 AI Concepts": {
"MoE (Mixture of Experts) 🧠": [
"What are Multi Agent Systems for Health",
"What is Mixture of Experts for Health",
"What are Semantic and Episodic Memory and what is Mirroring for Behavioral Health",
"What are Self Rewarding AI Systems for Health",
"How are AGI and AMI systems created using Multi Agent Systems and Mixture of Experts for Health"
],
"Multi Agent Systems (MAS) 🤝": [
"Distributed AI systems",
"Autonomous agents interacting",
"Cooperative and competitive behavior",
"Decentralized problem-solving",
"Applications in robotics, simulations, and more"
],
"Self Rewarding AI 🎁": [
"Intrinsic motivation for AI agents",
"Autonomous goal setting and achievement",
"Exploration and curiosity-driven learning",
"Potential for open-ended development",
"Research area in reinforcement learning"
],
"Semantic and Episodic Memory 📚": [
"Two types of long-term memory",
"Semantic: facts and general knowledge",
"Episodic: personal experiences and events",
"Crucial for AI systems to understand and reason",
"Research in knowledge representation and retrieval"
]
},
"🛠️ AI Tools & Platforms": {
"AutoGen 🔧": [
"Automated machine learning (AutoML) tool",
"Generates AI models based on requirements",
"Simplifies AI development process",
"Accessible to non-experts",
"Integration with various data sources"
],
"ChatDev 💬": [
"Platform for building chatbots and conversational AI",
"Drag-and-drop interface for designing chat flows",
"Pre-built templates and integrations",
"Supports multiple messaging platforms",
"Analytics and performance tracking"
],
"Omniverse 🌐": [
"Nvidia's 3D simulation and collaboration platform",
"Physically accurate virtual worlds",
"Supports AI training and testing",
"Used in industries like robotics, architecture, and gaming",
"Enables seamless collaboration and data exchange"
],
"Lumiere 🎥": [
"AI-powered video analytics platform",
"Extracts insights and metadata from video content",
"Facial recognition and object detection",
"Sentiment analysis and scene understanding",
"Applications in security, media, and marketing"
],
"SORA 🏗️": [
"Scalable Open Research Architecture",
"Framework for distributed AI research and development",
"Modular and extensible design",
"Facilitates collaboration and reproducibility",
"Supports various AI algorithms and models"
]
},
"🔬 Science Topics": {
"Physics 🔭": [
"Astrophysics: galaxies, cosmology, planets, high energy phenomena, instrumentation, solar/stellar",
"Condensed Matter: disordered systems, materials science, nano/mesoscale, quantum gases, soft matter, statistical mechanics, superconductivity",
"General Relativity and Quantum Cosmology",
"High Energy Physics: experiment, lattice, phenomenology, theory",
"Mathematical Physics",
"Nonlinear Sciences: adaptation, cellular automata, chaos, solvable systems, pattern formation",
"Nuclear: experiment, theory",
"Physics: accelerators, atmospherics, atomic/molecular, biophysics, chemical, computational, education, fluids, geophysics, optics, plasma, popular, space"
],
"Mathematics ➗": [
"Algebra: geometry, topology, number theory, combinatorics, representation theory",
"Analysis: PDEs, functional, numerical, spectral theory, ODEs, complex variables",
"Geometry: algebraic, differential, metric, symplectic, topological",
"Probability and Statistics",
"Applied Math: information theory, optimization and control"
],
"Computer Science 💻": [
"Artificial Intelligence and Machine Learning",
"Computation and Language, Complexity, Engineering, Finance, Science",
"Computer Vision, Graphics, Robotics",
"Cryptography, Security, Blockchain",
"Data Structures, Algorithms, Databases",
"Distributed and Parallel Computing",
"Formal Languages, Automata, Logic",
"Information Theory, Signal Processing",
"Networks, Internet Architecture, Social Networks",
"Programming Languages, Software Engineering"
],
"Quantitative Biology 🧬": [
"Biomolecules, Cell Behavior, Genomics",
"Molecular Networks, Neurons and Cognition",
"Populations, Evolution, Ecology",
"Quantitative Methods, Subcellular Processes",
"Tissues, Organs, Organisms"
],
"Quantitative Finance 📈": [
"Computational and Mathematical Finance",
"Econometrics and Statistical Finance",
"Economics, Portfolio Management, Trading",
"Pricing, Risk Management"
],
"Electrical Engineering 🔌": [
"Audio, Speech, Image and Video Processing",
"Communications and Information Theory",
"Signal Processing, Controls, Robotics",
"Electronic Circuits, Embedded Systems"
]
}
}
# This displays per video and per image.
@st.cache_resource
def display_glossary_entity(k):
search_urls = {
"🚀🌌ArXiv": lambda k: f"/?q={quote(k)}", # this url plus query!
"🃏Analyst": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix)}", # this url plus query!
"📚PyCoder": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix2)}", # this url plus query!
"🔬JSCoder": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix3)}", # this url plus query!
"📖Wiki": lambda k: f"https://en.wikipedia.org/wiki/{quote(k)}",
"🔍Google": lambda k: f"https://www.google.com/search?q={quote(k)}",
"🔎Bing": lambda k: f"https://www.bing.com/search?q={quote(k)}",
"🎥YouTube": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
"🐦Twitter": lambda k: f"https://twitter.com/search?q={quote(k)}",
}
links_md = ' '.join([f"[{emoji}]({url(k)})" for emoji, url in search_urls.items()])
#st.markdown(f"{k} {links_md}", unsafe_allow_html=True)
st.markdown(f"**{k}** <small>{links_md}</small>", unsafe_allow_html=True)
# Function to display the entire glossary in a grid format with links
@st.cache_resource
def display_glossary_grid(roleplaying_glossary):
search_urls = {
"🚀🌌ArXiv": lambda k: f"/?q={quote(k)}", # this url plus query!
"🃏Analyst": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix)}", # this url plus query!
"📚PyCoder": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix2)}", # this url plus query!
"🔬JSCoder": lambda k: f"/?q={quote(k)}-{quote(PromptPrefix3)}", # this url plus query!
"📖Wiki": lambda k: f"https://en.wikipedia.org/wiki/{quote(k)}",
"🔍Google": lambda k: f"https://www.google.com/search?q={quote(k)}",
"▶️YouTube": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
"🔎Bing": lambda k: f"https://www.bing.com/search?q={quote(k)}",
"🎥YouTube": lambda k: f"https://www.youtube.com/results?search_query={quote(k)}",
"🐦Twitter": lambda k: f"https://twitter.com/search?q={quote(k)}",
}
for category, details in roleplaying_glossary.items():
st.write(f"### {category}")
cols = st.columns(len(details)) # Create dynamic columns based on the number of games
#cols = st.columns(num_columns_text) # Create dynamic columns based on the number of games
for idx, (game, terms) in enumerate(details.items()):
with cols[idx]:
st.markdown(f"#### {game}")
for term in terms:
links_md = ' '.join([f"[{emoji}]({url(term)})" for emoji, url in search_urls.items()])
st.markdown(f"**{term}** <small>{links_md}</small>", unsafe_allow_html=True)
@st.cache_resource
def get_table_download_link(file_path):
try:
#with open(file_path, 'r') as file:
#with open(file_path, 'r', encoding="unicode", errors="surrogateescape") as file:
with open(file_path, 'r', encoding='utf-8') as file:
data = file.read()
b64 = base64.b64encode(data.encode()).decode()
file_name = os.path.basename(file_path)
ext = os.path.splitext(file_name)[1] # get the file extension
if ext == '.txt':
mime_type = 'text/plain'
elif ext == '.py':
mime_type = 'text/plain'
elif ext == '.xlsx':
mime_type = 'text/plain'
elif ext == '.csv':
mime_type = 'text/plain'
elif ext == '.htm':
mime_type = 'text/html'
elif ext == '.md':
mime_type = 'text/markdown'
elif ext == '.wav':
mime_type = 'audio/wav'
else:
mime_type = 'application/octet-stream' # general binary data type
href = f'<a href="data:{mime_type};base64,{b64}" target="_blank" download="{file_name}">{file_name}</a>'
return href
except:
return ''
@st.cache_resource
def create_zip_of_files(files): # ----------------------------------
zip_name = "Arxiv-Paper-Search-QA-RAG-Streamlit-Gradio-AP.zip"
with zipfile.ZipFile(zip_name, 'w') as zipf:
for file in files:
zipf.write(file)
return zip_name
@st.cache_resource
def get_zip_download_link(zip_file):
with open(zip_file, 'rb') as f:
data = f.read()
b64 = base64.b64encode(data).decode()
href = f'<a href="data:application/zip;base64,{b64}" download="{zip_file}">Download All</a>'
return href # ----------------------------------
def get_file():
st.write(st.session_state['file'])
def SaveFileTextClicked():
fileText = st.session_state.file_content_area
fileName = st.session_state.file_name_input
with open(fileName, 'w', encoding='utf-8') as file:
file.write(fileText)
st.markdown('Saved ' + fileName + '.')
def SaveFileNameClicked():
newFileName = st.session_state.file_name_input
oldFileName = st.session_state.filename
if (newFileName!=oldFileName):
os.rename(oldFileName, newFileName)
st.markdown('Renamed file ' + oldFileName + ' to ' + newFileName + '.')
newFileText = st.session_state.file_content_area
oldFileText = st.session_state.filetext
# Function to compare file sizes and delete duplicates
def compare_and_delete_files(files):
if not files:
st.warning("No files to compare.")
return
# Dictionary to store file sizes and their paths
file_sizes = {}
for file in files:
size = os.path.getsize(file)
if size in file_sizes:
file_sizes[size].append(file)
else:
file_sizes[size] = [file]
# Remove all but the latest file for each size group
for size, paths in file_sizes.items():
if len(paths) > 1:
latest_file = max(paths, key=os.path.getmtime)
for file in paths:
if file != latest_file:
os.remove(file)
st.success(f"Deleted {file} as a duplicate.")
st.rerun()
# Function to get file size
def get_file_size(file_path):
return os.path.getsize(file_path)
def FileSidebar():
# File Sidebar for files 🌐View, 📂Open, ▶️Run, and 🗑Delete per file
all_files = glob.glob("*.md")
all_files = [file for file in all_files if len(os.path.splitext(file)[0]) >= 10] # exclude files with short names
all_files.sort(key=lambda x: (os.path.splitext(x)[1], x), reverse=True) # sort by filename length which puts similar prompts together - consider making date and time of file optional.
# Button to compare files and delete duplicates
#if st.button("Compare and Delete Duplicates"):
# compare_and_delete_files(all_files)
# ⬇️ Download
Files1, Files2 = st.sidebar.columns(2)
with Files1:
if st.button("🗑 Delete All"):
for file in all_files:
os.remove(file)
st.rerun()
with Files2:
if st.button("⬇️ Download"):
zip_file = create_zip_of_files(all_files)
st.sidebar.markdown(get_zip_download_link(zip_file), unsafe_allow_html=True)
file_contents=''
file_name=''
next_action=''
# Add files 🌐View, 📂Open, ▶️Run, and 🗑Delete per file
for file in all_files:
col1, col2, col3, col4, col5 = st.sidebar.columns([1,6,1,1,1]) # adjust the ratio as needed
with col1:
if st.button("🌐", key="md_"+file): # md emoji button
file_contents = load_file(file)
file_name=file
next_action='md'
st.session_state['next_action'] = next_action
with col2:
st.markdown(get_table_download_link(file), unsafe_allow_html=True)
with col3:
if st.button("📂", key="open_"+file): # open emoji button
file_contents = load_file(file)
file_name=file
next_action='open'
st.session_state['lastfilename'] = file
st.session_state['filename'] = file
st.session_state['filetext'] = file_contents
st.session_state['next_action'] = next_action
with col4:
if st.button("▶️", key="read_"+file): # search emoji button
file_contents = load_file(file)
file_name=file
next_action='search'
st.session_state['next_action'] = next_action
with col5:
if st.button("🗑", key="delete_"+file):
os.remove(file)
file_name=file
st.rerun()
next_action='delete'
st.session_state['next_action'] = next_action
# 🚩File duplicate detector - useful to prune and view all. Pruning works well by file size detection of two similar and flags the duplicate.
file_sizes = [get_file_size(file) for file in all_files]
previous_size = None
st.sidebar.title("File Operations")
for file, size in zip(all_files, file_sizes):
duplicate_flag = "🚩" if size == previous_size else ""
with st.sidebar.expander(f"File: {file} {duplicate_flag}"):
st.text(f"Size: {size} bytes")
if st.button("View", key=f"view_{file}"):
try:
with open(file, "r", encoding='utf-8') as f: # Ensure the file is read with UTF-8 encoding
file_content = f.read()
st.code(file_content, language="markdown")
except UnicodeDecodeError:
st.error("Failed to decode the file with UTF-8. It might contain non-UTF-8 encoded characters.")
if st.button("Delete", key=f"delete3_{file}"):
os.remove(file)
st.rerun()
previous_size = size # Update previous size for the next iteration
if len(file_contents) > 0:
if next_action=='open': # For "open", prep session state if it hasn't been yet
if 'lastfilename' not in st.session_state:
st.session_state['lastfilename'] = ''
if 'filename' not in st.session_state:
st.session_state['filename'] = ''
if 'filetext' not in st.session_state:
st.session_state['filetext'] = ''
open1, open2 = st.columns(spec=[.8,.2])
with open1:
# Use onchange functions to autoexecute file name and text save functions.
file_name_input = st.text_input(key='file_name_input', on_change=SaveFileNameClicked, label="File Name:",value=file_name )
file_content_area = st.text_area(key='file_content_area', on_change=SaveFileTextClicked, label="File Contents:", value=file_contents, height=300)
ShowButtons = False # Having buttons is redundant. They work but if on change event seals the deal so be it - faster save is less impedence - less context breaking
if ShowButtons:
bp1,bp2 = st.columns([.5,.5])
with bp1:
if st.button(label='💾 Save Name'):
SaveFileNameClicked()
with bp2:
if st.button(label='💾 Save File'):
SaveFileTextClicked()
new_file_content_area = st.session_state['file_content_area']
if new_file_content_area != file_contents:
st.markdown(new_file_content_area) #changed
if st.button("🔍 Run AI Meta Strategy", key="filecontentssearch"):
#search_glossary(file_content_area)
filesearch = PromptPrefix + file_content_area
st.markdown(filesearch)
if st.button(key=rerun, label='🔍AI Search' ):
search_glossary(filesearch)
if next_action=='md':
st.markdown(file_contents)
buttonlabel = '🔍Run'
if st.button(key='Runmd', label = buttonlabel):
user_prompt = file_contents
#try:
search_glossary(file_contents)
#except:
#st.markdown('GPT is sleeping. Restart ETA 30 seconds.')
if next_action=='search':
file_content_area = st.text_area("File Contents:", file_contents, height=500)
user_prompt = file_contents
#try:
#search_glossary(file_contents)
filesearch = PromptPrefix2 + file_content_area
st.markdown(filesearch)
if st.button(key=rerun, label='🔍Re-Code' ):
search_glossary(filesearch)
#except:
#st.markdown('GPT is sleeping. Restart ETA 30 seconds.')
# ----------------------------------------------------- File Sidebar for Jump Gates ------------------------------------------
# Randomly select a title
titles = [
"🎺🎷 The Sounds 🎹🥁 of the Big Easy 🎭🎉",
"🎼🎸 NOLA's Iconic 🎤🪕 Musical 🔊 Heritage 🏰",
"🎺🪘 Crescent City 🌙 Rhythms & Grooves 🎹💃",
"🎷🎸 Mardi Gras 🎭 Melodies",
"🎼🎺 Straight Outta Nawlins ⚜️",
"🥁🎻 Jazzy 🎷 Jambalaya 🍛 of New Orleans",
"🏰 Musical 🎹 Soul 🙌",
"🥁🎻 The Music Of New Orleans MoE 🎭🎉"
]
selected_title = random.choice(titles)
st.markdown(f"**{selected_title}**")
FileSidebar()
# ---- Art Card Sidebar with Random Selection of image:
def get_image_as_base64(url):
response = requests.get(url)
if response.status_code == 200:
# Convert the image to base64
return base64.b64encode(response.content).decode("utf-8")
else:
return None
def create_download_link(filename, base64_str):
href = f'<a href="data:file/png;base64,{base64_str}" download="{filename}">Download Image</a>'
return href
@st.cache_resource
def SideBarImageShuffle():
image_urls = [
"https://cdn-uploads.huggingface.co/production/uploads/620630b603825909dcbeba35/cfhJIasuxLkT5fnaAE6Gj.png",
"https://cdn-uploads.huggingface.co/production/uploads/620630b603825909dcbeba35/UMo4oWNrrd6RLLzsFxQAi.png",
"https://cdn-uploads.huggingface.co/production/uploads/620630b603825909dcbeba35/o_EH4cTs5Qxiu7xTZw9I3.png",
"https://cdn-uploads.huggingface.co/production/uploads/620630b603825909dcbeba35/cmCZ5RTdSx3usMm7MwwWK.png",
]
selected_image_url = random.choice(image_urls)
selected_image_base64 = get_image_as_base64(selected_image_url)
if selected_image_base64 is not None:
with st.sidebar:
st.markdown(f"")
else:
st.sidebar.write("Failed to load the image.")
ShowSideImages=False
if ShowSideImages:
SideBarImageShuffle()
# Ensure the directory for storing scores exists
score_dir = "scores"
os.makedirs(score_dir, exist_ok=True)
# Function to generate a unique key for each button, including an emoji
def generate_key(label, header, idx):
return f"{header}_{label}_{idx}_key"
# Function to increment and save score
def update_score(key, increment=1):
score_file = os.path.join(score_dir, f"{key}.json")
if os.path.exists(score_file):
with open(score_file, "r") as file:
score_data = json.load(file)
else:
score_data = {"clicks": 0, "score": 0}
score_data["clicks"] += 1
score_data["score"] += increment
with open(score_file, "w") as file:
json.dump(score_data, file)
return score_data["score"]
# Function to load score
def load_score(key):
score_file = os.path.join(score_dir, f"{key}.json")
if os.path.exists(score_file):
with open(score_file, "r") as file:
score_data = json.load(file)
return score_data["score"]
return 0
# 🔍Run--------------------------------------------------------
@st.cache_resource
def search_glossary(query):
#for category, terms in roleplaying_glossary.items():
# if query.lower() in (term.lower() for term in terms):
# st.markdown(f"#### {category}")
# st.write(f"- {query}")
all=""
st.markdown(f"- {query}")
# 🔍Run 1 - plain query
#response = chat_with_model(query)
#response1 = chat_with_model45(query)
#all = query + ' ' + response1
#st.write('🔍Run 1 is Complete.')
# ArXiv searcher ~-<>-~ Paper Summary - Ask LLM
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
response2 = client.predict(
query, # str in 'parameter_13' Textbox component
"mistralai/Mixtral-8x7B-Instruct-v0.1", # Literal['mistralai/Mixtral-8x7B-Instruct-v0.1', 'mistralai/Mistral-7B-Instruct-v0.2', 'google/gemma-7b-it', 'None'] in 'LLM Model' Dropdown component
True, # bool in 'Stream output' Checkbox component
api_name="/ask_llm"
)
st.write('🔍Run of Multi-Agent System Paper Summary Spec is Complete')
st.markdown(response2)
# ArXiv searcher ~-<>-~ Paper References - Update with RAG
client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
response1 = client.predict(
query,
10,
"Semantic Search - up to 10 Mar 2024", # Literal['Semantic Search - up to 10 Mar 2024', 'Arxiv Search - Latest - (EXPERIMENTAL)'] in 'Search Source' Dropdown component
"mistralai/Mixtral-8x7B-Instruct-v0.1", # Literal['mistralai/Mixtral-8x7B-Instruct-v0.1', 'mistralai/Mistral-7B-Instruct-v0.2', 'google/gemma-7b-it', 'None'] in 'LLM Model' Dropdown component
api_name="/update_with_rag_md"
)
st.write('🔍Run of Multi-Agent System Paper References is Complete')
#st.markdown(response1)
responseall = response2 + response1[0] + response1[1]
st.markdown(responseall)
return responseall
# GPT 35 turbo and GPT 45 - - - - - - - - - - - - -<><><><><>:
RunPostArxivLLM = False
if RunPostArxivLLM:
# 🔍Run PaperSummarizer
PaperSummarizer = ' Create a paper summary as a markdown table with paper links clustering the features writing short markdown emoji outlines to extract three main ideas from each of the ten summaries. For each one create three simple points led by an emoji of the main three steps needed as method step process for implementing the idea as a single app.py streamlit python app. '
response2 = chat_with_model(PaperSummarizer + str(response1))
st.write('🔍Run 3 - Paper Summarizer is Complete.')
# 🔍Run AppSpecifier
AppSpecifier = ' Design and write a streamlit python code listing and specification that implements each scientific method steps as ten functions keeping specification in a markdown table in the function comments with original paper link to outline the AI pipeline ensemble implementing code as full plan to build.'
response3 = chat_with_model(AppSpecifier + str(response2))
st.write('🔍Run 4 - AppSpecifier is Complete.')
# 🔍Run PythonAppCoder
PythonAppCoder = ' Complete this streamlit python app implementing the functions in detail using appropriate python libraries and streamlit user interface elements. Show full code listing for the completed detail app as full code listing with no comments or commentary. '
#result = str(result).replace('\n', ' ').replace('|', ' ')
# response4 = chat_with_model45(PythonAppCoder + str(response3))
response4 = chat_with_model(PythonAppCoder + str(response3))
st.write('🔍Run Python AppCoder is Complete.')
# experimental 45 - - - - - - - - - - - - -<><><><><>
responseAll = '# Query: ' + query + '# Summary: ' + str(response2) + '# Streamlit App Specifier: ' + str(response3) + '# Complete Streamlit App: ' + str(response4) + '# Scholarly Article Links References: ' + str(response1)
filename = generate_filename(responseAll, "md")
create_file(filename, query, responseAll, should_save)
return responseAll # 🔍Run--------------------------------------------------------
else:
return response1
# Function to display the glossary in a structured format
def display_glossary(glossary, area):
if area in glossary:
st.subheader(f"📘 Glossary for {area}")
for game, terms in glossary[area].items():
st.markdown(f"### {game}")
for idx, term in enumerate(terms, start=1):
st.write(f"{idx}. {term}")
#@st.cache_resource
def display_videos_and_links(num_columns):
video_files = [f for f in os.listdir('.') if f.endswith('.mp4')]
if not video_files:
st.write("No MP4 videos found in the current directory.")
return
video_files_sorted = sorted(video_files, key=lambda x: len(x.split('.')[0]))
cols = st.columns(num_columns) # Define num_columns columns outside the loop
col_index = 0 # Initialize column index
for video_file in video_files_sorted:
with cols[col_index % num_columns]: # Use modulo 2 to alternate between the first and second column
# Embedding video with autoplay and loop using HTML
#video_html = ("""<video width="100%" loop autoplay> <source src="{video_file}" type="video/mp4">Your browser does not support the video tag.</video>""")
#st.markdown(video_html, unsafe_allow_html=True)
k = video_file.split('.')[0] # Assumes keyword is the file name without extension
st.video(video_file, format='video/mp4', start_time=0)
display_glossary_entity(k)
col_index += 1 # Increment column index to place the next video in the next column
@st.cache_resource
def display_images_and_wikipedia_summaries(num_columns=4):
image_files = [f for f in os.listdir('.') if f.endswith('.png')]
if not image_files:
st.write("No PNG images found in the current directory.")
return
image_files_sorted = sorted(image_files, key=lambda x: len(x.split('.')[0]))
cols = st.columns(num_columns) # Use specified num_columns for layout
col_index = 0 # Initialize column index for cycling through columns
for image_file in image_files_sorted:
with cols[col_index % num_columns]: # Cycle through columns based on num_columns
image = Image.open(image_file)
st.image(image, caption=image_file, use_column_width=True)
k = image_file.split('.')[0] # Assumes keyword is the file name without extension
display_glossary_entity(k)
col_index += 1 # Increment to move to the next column in the next iteration
def get_all_query_params(key):
return st.query_params().get(key, [])
def clear_query_params():
st.query_params()
# Function to display content or image based on a query
#@st.cache_resource
def display_content_or_image(query):
for category, terms in transhuman_glossary.items():
for term in terms:
if query.lower() in term.lower():
st.subheader(f"Found in {category}:")
st.write(term)
return True # Return after finding and displaying the first match
image_dir = "images" # Example directory where images are stored
image_path = f"{image_dir}/{query}.png" # Construct image path with query
if os.path.exists(image_path):
st.image(image_path, caption=f"Image for {query}")
return True
st.warning("No matching content or image found.")
return False
game_emojis = {
"Dungeons and Dragons": "🐉",
"Call of Cthulhu": "🐙",
"GURPS": "🎲",
"Pathfinder": "🗺️",
"Kindred of the East": "🌅",
"Changeling": "🍃",
}
topic_emojis = {
"Core Rulebooks": "📚",
"Maps & Settings": "🗺️",
"Game Mechanics & Tools": "⚙️",
"Monsters & Adversaries": "👹",
"Campaigns & Adventures": "📜",
"Creatives & Assets": "🎨",
"Game Master Resources": "🛠️",
"Lore & Background": "📖",
"Character Development": "🧍",
"Homebrew Content": "🔧",
"General Topics": "🌍",
}
# Adjusted display_buttons_with_scores function
def display_buttons_with_scores(num_columns_text):
for category, games in roleplaying_glossary.items():
category_emoji = topic_emojis.get(category, "🔍") # Default to search icon if no match
st.markdown(f"## {category_emoji} {category}")
for game, terms in games.items():
game_emoji = game_emojis.get(game, "🎮") # Default to generic game controller if no match
for term in terms:
key = f"{category}_{game}_{term}".replace(' ', '_').lower()
score = load_score(key)
if st.button(f"{game_emoji} {category} {game} {term} {score}", key=key):
update_score(key)
# Create a dynamic query incorporating emojis and formatting for clarity
query_prefix = f"{category_emoji} {game_emoji} ** {category} - {game} - {term} - **"
# ----------------------------------------------------------------------------------------------
#query_body = f"Create a detailed outline for **{term}** with subpoints highlighting key aspects, using emojis for visual engagement. Include step-by-step rules and boldface important entities and ruleset elements."
query_body = f"Create a streamlit python app.py that produces a detailed markdown outline and emoji laden user interface with labels with the entity name and emojis in all labels with a set of streamlit UI components with drop down lists and dataframes and buttons with expander and sidebar for the app to run the data as default values mostly in text boxes. Feature a 3 point outline sith 3 subpoints each where each line has about six words describing this and also contain appropriate emoji for creating sumamry of all aspeccts of this topic. an outline for **{term}** with subpoints highlighting key aspects, using emojis for visual engagement. Include step-by-step rules and boldface important entities and ruleset elements."
response = search_glossary(query_prefix + query_body)
def get_all_query_params(key):
return st.query_params().get(key, [])
def clear_query_params():
st.query_params()
# My Inference API Copy
API_URL = 'https://qe55p8afio98s0u3.us-east-1.aws.endpoints.huggingface.cloud' # Dr Llama
# Meta's Original - Chat HF Free Version:
#API_URL = "https://api-inference.huggingface.co/models/meta-llama/Llama-2-7b-chat-hf"
API_KEY = os.getenv('API_KEY')
MODEL1="meta-llama/Llama-2-7b-chat-hf"
MODEL1URL="https://huggingface.co/meta-llama/Llama-2-7b-chat-hf"
HF_KEY = os.getenv('HF_KEY')
headers = {
"Authorization": f"Bearer {HF_KEY}",
"Content-Type": "application/json"
}
key = os.getenv('OPENAI_API_KEY')
prompt = "...."
should_save = st.sidebar.checkbox("💾 Save", value=True, help="Save your session data.")
# 3. Stream Llama Response
@st.cache_resource
def StreamLLMChatResponse(prompt):
try:
endpoint_url = API_URL
hf_token = API_KEY
st.write('Running client ' + endpoint_url)
client = InferenceClient(endpoint_url, token=hf_token)
gen_kwargs = dict(
max_new_tokens=512,
top_k=30,
top_p=0.9,
temperature=0.2,
repetition_penalty=1.02,
stop_sequences=["\nUser:", "<|endoftext|>", "</s>"],
)
stream = client.text_generation(prompt, stream=True, details=True, **gen_kwargs)
report=[]
res_box = st.empty()
collected_chunks=[]
collected_messages=[]
allresults=''
for r in stream:
if r.token.special:
continue
if r.token.text in gen_kwargs["stop_sequences"]:
break
collected_chunks.append(r.token.text)
chunk_message = r.token.text
collected_messages.append(chunk_message)
try:
report.append(r.token.text)
if len(r.token.text) > 0:
result="".join(report).strip()
res_box.markdown(f'*{result}*')
except:
st.write('Stream llm issue')
SpeechSynthesis(result)
return result
except:
st.write('Llama model is asleep. Starting up now on A10 - please give 5 minutes then retry as KEDA scales up from zero to activate running container(s).')
# 4. Run query with payload
def query(payload):
response = requests.post(API_URL, headers=headers, json=payload)
st.markdown(response.json())
return response.json()
def get_output(prompt):
return query({"inputs": prompt})
# 5. Auto name generated output files from time and content
def generate_filename(prompt, file_type):
central = pytz.timezone('US/Central')
safe_date_time = datetime.now(central).strftime("%m%d_%H%M")
replaced_prompt = prompt.replace(" ", "_").replace("\n", "_")
safe_prompt = "".join(x for x in replaced_prompt if x.isalnum() or x == "_")[:255] # 255 is linux max, 260 is windows max
#safe_prompt = "".join(x for x in replaced_prompt if x.isalnum() or x == "_")[:45]
return f"{safe_date_time}_{safe_prompt}.{file_type}"
# 6. Speech transcription via OpenAI service
def transcribe_audio(openai_key, file_path, model):
openai.api_key = openai_key
OPENAI_API_URL = "https://api.openai.com/v1/audio/transcriptions"
headers = {
"Authorization": f"Bearer {openai_key}",
}
with open(file_path, 'rb') as f:
data = {'file': f}
st.write('STT transcript ' + OPENAI_API_URL)
response = requests.post(OPENAI_API_URL, headers=headers, files=data, data={'model': model})
if response.status_code == 200:
st.write(response.json())
chatResponse = chat_with_model(response.json().get('text'), '') # *************************************
transcript = response.json().get('text')
filename = generate_filename(transcript, 'txt')
response = chatResponse
user_prompt = transcript
create_file(filename, user_prompt, response, should_save)
return transcript
else:
st.write(response.json())
st.error("Error in API call.")
return None
# 7. Auto stop on silence audio control for recording WAV files
def save_and_play_audio(audio_recorder):
audio_bytes = audio_recorder(key='audio_recorder')
if audio_bytes:
filename = generate_filename("Recording", "wav")
with open(filename, 'wb') as f:
f.write(audio_bytes)
st.audio(audio_bytes, format="audio/wav")
return filename
return None
# 8. File creator that interprets type and creates output file for text, markdown and code
def create_file(filename, prompt, response, should_save=True):
if not should_save:
return
base_filename, ext = os.path.splitext(filename)
if ext in ['.txt', '.htm', '.md']:
# ****** line 344 is read utf-8 encoding was needed when running locally to save utf-8 encoding and not fail on write
#with open(f"{base_filename}.md", 'w') as file:
#with open(f"{base_filename}.md", 'w', encoding="ascii", errors="surrogateescape") as file:
with open(f"{base_filename}.md", 'w', encoding='utf-8') as file:
#try:
#content = (prompt.strip() + '\r\n' + decode(response, ))
file.write(response)
#except:
# st.write('.')
# ****** utf-8 encoding was needed when running locally to save utf-8 encoding and not fail on write
#has_python_code = re.search(r"```python([\s\S]*?)```", prompt.strip() + '\r\n' + response)
#has_python_code = bool(re.search(r"```python([\s\S]*?)```", prompt.strip() + '\r\n' + response))
#if has_python_code:
# python_code = re.findall(r"```python([\s\S]*?)```", response)[0].strip()
# with open(f"{base_filename}-Code.py", 'w') as file:
# file.write(python_code)
# with open(f"{base_filename}.md", 'w') as file:
# content = prompt.strip() + '\r\n' + response
# file.write(content)
def truncate_document(document, length):
return document[:length]
def divide_document(document, max_length):
return [document[i:i+max_length] for i in range(0, len(document), max_length)]
def CompressXML(xml_text):
root = ET.fromstring(xml_text)
for elem in list(root.iter()):
if isinstance(elem.tag, str) and 'Comment' in elem.tag:
elem.parent.remove(elem)
return ET.tostring(root, encoding='unicode', method="xml")
# 10. Read in and provide UI for past files
@st.cache_resource
def read_file_content(file,max_length):
if file.type == "application/json":
content = json.load(file)
return str(content)
elif file.type == "text/html" or file.type == "text/htm":
content = BeautifulSoup(file, "html.parser")
return content.text
elif file.type == "application/xml" or file.type == "text/xml":
tree = ET.parse(file)
root = tree.getroot()
xml = CompressXML(ET.tostring(root, encoding='unicode'))
return xml
elif file.type == "text/markdown" or file.type == "text/md":
md = mistune.create_markdown()
content = md(file.read().decode())
return content
elif file.type == "text/plain":
return file.getvalue().decode()
else:
return ""
# 11. Chat with GPT - Caution on quota - now favoring fastest AI pipeline STT Whisper->LLM Llama->TTS
@st.cache_resource
def chat_with_model(prompt, document_section='', model_choice='gpt-3.5-turbo'): # gpt-4-0125-preview gpt-3.5-turbo
model = model_choice
conversation = [{'role': 'system', 'content': 'You are a coder, inventor, and writer of quotes on wisdom as a helpful expert in all fields of health, math, development and AI using python.'}]
conversation.append({'role': 'user', 'content': prompt})
if len(document_section)>0:
conversation.append({'role': 'assistant', 'content': document_section})
start_time = time.time()
report = []
res_box = st.empty()
collected_chunks = []
collected_messages = []
for chunk in openai.ChatCompletion.create(model=model_choice, messages=conversation, temperature=0.5, stream=True):
collected_chunks.append(chunk)
chunk_message = chunk['choices'][0]['delta']
collected_messages.append(chunk_message)
content=chunk["choices"][0].get("delta",{}).get("content")
try:
report.append(content)
if len(content) > 0:
result = "".join(report).strip()
res_box.markdown(f'*{result}*')
except:
st.write(' ')
full_reply_content = ''.join([m.get('content', '') for m in collected_messages])
st.write("Elapsed time:")
st.write(time.time() - start_time)
return full_reply_content
# 11.1 45
@st.cache_resource
def chat_with_model45(prompt, document_section='', model_choice='gpt-4-0125-preview'): # gpt-4-0125-preview gpt-3.5-turbo
model = model_choice
conversation = [{'role': 'system', 'content': 'You are a coder, inventor, and writer of quotes on wisdom as a helpful expert in all fields of health, math, development and AI using python.'}]
conversation.append({'role': 'user', 'content': prompt})
if len(document_section)>0:
conversation.append({'role': 'assistant', 'content': document_section})
start_time = time.time()
report = []
res_box = st.empty()
collected_chunks = []
collected_messages = []
for chunk in openai.ChatCompletion.create(model=model_choice, messages=conversation, temperature=0.5, stream=True):
collected_chunks.append(chunk)
chunk_message = chunk['choices'][0]['delta']
collected_messages.append(chunk_message)
content=chunk["choices"][0].get("delta",{}).get("content")
try:
report.append(content)
if len(content) > 0:
result = "".join(report).strip()
res_box.markdown(f'*{result}*')
except:
st.write(' ')
full_reply_content = ''.join([m.get('content', '') for m in collected_messages])
st.write("Elapsed time:")
st.write(time.time() - start_time)
return full_reply_content
@st.cache_resource
def chat_with_file_contents(prompt, file_content, model_choice='gpt-3.5-turbo'): # gpt-4-0125-preview gpt-3.5-turbo
#def chat_with_file_contents(prompt, file_content, model_choice='gpt-4-0125-preview'): # gpt-4-0125-preview gpt-3.5-turbo
conversation = [{'role': 'system', 'content': 'You are a helpful assistant.'}]
conversation.append({'role': 'user', 'content': prompt})
if len(file_content)>0:
conversation.append({'role': 'assistant', 'content': file_content})
response = openai.ChatCompletion.create(model=model_choice, messages=conversation)
return response['choices'][0]['message']['content']
def extract_mime_type(file):
if isinstance(file, str):
pattern = r"type='(.*?)'"
match = re.search(pattern, file)
if match:
return match.group(1)
else:
raise ValueError(f"Unable to extract MIME type from {file}")
elif isinstance(file, streamlit.UploadedFile):
return file.type
else:
raise TypeError("Input should be a string or a streamlit.UploadedFile object")
def extract_file_extension(file):
# get the file name directly from the UploadedFile object
file_name = file.name
pattern = r".*?\.(.*?)$"
match = re.search(pattern, file_name)
if match:
return match.group(1)
else:
raise ValueError(f"Unable to extract file extension from {file_name}")
# Normalize input as text from PDF and other formats
@st.cache_resource
def pdf2txt(docs):
text = ""
for file in docs:
file_extension = extract_file_extension(file)
st.write(f"File type extension: {file_extension}")
if file_extension.lower() in ['py', 'txt', 'html', 'htm', 'xml', 'json']:
text += file.getvalue().decode('utf-8')
elif file_extension.lower() == 'pdf':
from PyPDF2 import PdfReader
pdf = PdfReader(BytesIO(file.getvalue()))
for page in range(len(pdf.pages)):
text += pdf.pages[page].extract_text() # new PyPDF2 syntax
return text
def txt2chunks(text):
text_splitter = CharacterTextSplitter(separator="\n", chunk_size=1000, chunk_overlap=200, length_function=len)
return text_splitter.split_text(text)
# Vector Store using FAISS
@st.cache_resource
def vector_store(text_chunks):
embeddings = OpenAIEmbeddings(openai_api_key=key)
return FAISS.from_texts(texts=text_chunks, embedding=embeddings)
# Memory and Retrieval chains
@st.cache_resource
def get_chain(vectorstore):
llm = ChatOpenAI()
memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True)
return ConversationalRetrievalChain.from_llm(llm=llm, retriever=vectorstore.as_retriever(), memory=memory)
def process_user_input(user_question):
response = st.session_state.conversation({'question': user_question})
st.session_state.chat_history = response['chat_history']
for i, message in enumerate(st.session_state.chat_history):
template = user_template if i % 2 == 0 else bot_template
st.write(template.replace("{{MSG}}", message.content), unsafe_allow_html=True)
filename = generate_filename(user_question, 'txt')
response = message.content
user_prompt = user_question
create_file(filename, user_prompt, response, should_save)
def divide_prompt(prompt, max_length):
words = prompt.split()
chunks = []
current_chunk = []
current_length = 0
for word in words:
if len(word) + current_length <= max_length:
current_length += len(word) + 1
current_chunk.append(word)
else:
chunks.append(' '.join(current_chunk))
current_chunk = [word]
current_length = len(word)
chunks.append(' '.join(current_chunk))
return chunks
API_URL_IE = f'https://tonpixzfvq3791u9.us-east-1.aws.endpoints.huggingface.cloud'
API_URL_IE = "https://api-inference.huggingface.co/models/openai/whisper-small.en"
MODEL2 = "openai/whisper-small.en"
MODEL2_URL = "https://huggingface.co/openai/whisper-small.en"
HF_KEY = st.secrets['HF_KEY']
headers = {
"Authorization": f"Bearer {HF_KEY}",
"Content-Type": "audio/wav"
}
def query(filename):
with open(filename, "rb") as f:
data = f.read()
response = requests.post(API_URL_IE, headers=headers, data=data)
return response.json()
def generate_filename(prompt, file_type):
central = pytz.timezone('US/Central')
safe_date_time = datetime.now(central).strftime("%m%d_%H%M")
replaced_prompt = prompt.replace(" ", "_").replace("\n", "_")
safe_prompt = "".join(x for x in replaced_prompt if x.isalnum() or x == "_")[:90]
return f"{safe_date_time}_{safe_prompt}.{file_type}"
# 15. Audio recorder to Wav file
def save_and_play_audio(audio_recorder):
audio_bytes = audio_recorder()
if audio_bytes:
filename = generate_filename("Recording", "wav")
with open(filename, 'wb') as f:
f.write(audio_bytes)
st.audio(audio_bytes, format="audio/wav")
return filename
# 16. Speech transcription to file output
def transcribe_audio(filename):
output = query(filename)
return output
# Sample function to demonstrate a response, replace with your own logic
def StreamMedChatResponse(topic):
st.write(f"Showing resources or questions related to: {topic}")
# Function to encode file to base64
def get_base64_encoded_file(file_path):
with open(file_path, "rb") as file:
return base64.b64encode(file.read()).decode()
# Function to create a download link
def get_audio_download_link(file_path):
base64_file = get_base64_encoded_file(file_path)
return f'<a href="data:file/wav;base64,{base64_file}" download="{os.path.basename(file_path)}">⬇️ Download Audio</a>'
# 🎵 Wav Audio files - Transcription History in Wav
all_files = glob.glob("*.wav")
all_files = [file for file in all_files if len(os.path.splitext(file)[0]) >= 10] # exclude files with short names
all_files.sort(key=lambda x: (os.path.splitext(x)[1], x), reverse=True) # sort by file type and file name in descending order
filekey = 'delall'
if st.sidebar.button("🗑 Delete All Audio", key=filekey):
for file in all_files:
os.remove(file)
st.rerun()
for file in all_files:
col1, col2 = st.sidebar.columns([6, 1]) # adjust the ratio as needed
with col1:
st.markdown(file)
if st.button("🎵", key="play_" + file): # play emoji button
audio_file = open(file, 'rb')
audio_bytes = audio_file.read()
st.audio(audio_bytes, format='audio/wav')
#st.markdown(get_audio_download_link(file), unsafe_allow_html=True)
#st.text_input(label="", value=file)
with col2:
if st.button("🗑", key="delete_" + file):
os.remove(file)
st.rerun()
GiveFeedback=False
if GiveFeedback:
with st.expander("Give your feedback 👍", expanded=False):
feedback = st.radio("Step 8: Give your feedback", ("👍 Upvote", "👎 Downvote"))
if feedback == "👍 Upvote":
st.write("You upvoted 👍. Thank you for your feedback!")
else:
st.write("You downvoted 👎. Thank you for your feedback!")
load_dotenv()
st.write(css, unsafe_allow_html=True)
st.header("Chat with documents :books:")
user_question = st.text_input("Ask a question about your documents:")
if user_question:
process_user_input(user_question)
with st.sidebar:
st.subheader("Your documents")
docs = st.file_uploader("import documents", accept_multiple_files=True)
with st.spinner("Processing"):
raw = pdf2txt(docs)
if len(raw) > 0:
length = str(len(raw))
text_chunks = txt2chunks(raw)
vectorstore = vector_store(text_chunks)
st.session_state.conversation = get_chain(vectorstore)
st.markdown('# AI Search Index of Length:' + length + ' Created.') # add timing
filename = generate_filename(raw, 'txt')
create_file(filename, raw, '', should_save)
try:
query_params = st.query_params
query = (query_params.get('q') or query_params.get('query') or [''])
if len(query) > 1:
result = search_arxiv(query)
#result2 = search_glossary(result)
except:
st.markdown(' ')
if 'action' in st.query_params:
action = st.query_params()['action'][0] # Get the first (or only) 'action' parameter
if action == 'show_message':
st.success("Showing a message because 'action=show_message' was found in the URL.")
elif action == 'clear':
clear_query_params()
st.rerun()
if 'query' in st.query_params:
query = st.query_params['query'][0] # Get the query parameter
# Display content or image based on the query
display_content_or_image(query)
def transcribe_canary(filename):
from gradio_client import Client
client = Client("https://awacke1-speech-recognition-canary-nvidiat4.hf.space/")
result = client.predict(
filename, # filepath in 'parameter_5' Audio component
"English", # Literal['English', 'Spanish', 'French', 'German'] in 'Input audio is spoken in:' Dropdown component
"English", # Literal['English', 'Spanish', 'French', 'German'] in 'Transcribe in language:' Dropdown component
True, # bool in 'Punctuation & Capitalization in transcript?' Checkbox component
api_name="/transcribe"
)
st.write(result)
return result
filename = save_and_play_audio(audio_recorder)
if filename is not None:
transcript=''
transcript=transcribe_canary(filename)
result = search_arxiv(transcript)
#result2 = search_glossary(result)
#st.markdown(result)
#st.markdown
#transcription = transcribe_audio(filename)
#try:
# transcript = transcription['text']
# st.write(transcript)
#except:
# transcript=''
# st.write(transcript)
#st.write('Reasoning with your inputs..')
#response = chat_with_model(transcript)
#st.write('Response:')
#st.write(response)
#filename = generate_filename(response, "txt")
#create_file(filename, transcript, response, should_save)
# Whisper to Llama:
response = StreamLLMChatResponse(transcript)
filename_txt = generate_filename(transcript, "md")
create_file(filename_txt, transcript, response, should_save)
filename_wav = filename_txt.replace('.txt', '.wav')
import shutil
try:
if os.path.exists(filename):
shutil.copyfile(filename, filename_wav)
except:
st.write('.')
if os.path.exists(filename):
os.remove(filename)
prompt = '''
What is MoE?
What are Multi Agent Systems?
What is Self Rewarding AI?
What is Semantic and Episodic memory?
What is AutoGen?
What is ChatDev?
What is Omniverse?
What is Lumiere?
What is SORA?
'''
session_state = {}
if "search_queries" not in session_state:
session_state["search_queries"] = []
example_input = st.text_input("Search", value=session_state["search_queries"][-1] if session_state["search_queries"] else "")
if example_input:
session_state["search_queries"].append(example_input)
# Search AI
query=example_input
if query:
result = search_arxiv(query)
#search_glossary(query)
search_glossary(result)
st.markdown(' ')
#st.write("Search history:")
for example_input in session_state["search_queries"]:
st.write(example_input)
if st.button("Run Prompt", help="Click to run."):
try:
response=StreamLLMChatResponse(example_input)
create_file(filename, example_input, response, should_save)
except:
st.write('model is asleep. Starting now on A10 GPU. Please wait one minute then retry. KEDA triggered.')
openai.api_key = os.getenv('OPENAI_API_KEY')
if openai.api_key == None: openai.api_key = st.secrets['OPENAI_API_KEY']
menu = ["txt", "htm", "xlsx", "csv", "md", "py"]
choice = st.sidebar.selectbox("Output File Type:", menu)
#model_choice = st.sidebar.radio("Select Model:", ('gpt-3.5-turbo', 'gpt-3.5-turbo-0301'))
#user_prompt = st.text_area("Enter prompts, instructions & questions:", '', height=100)
AddAFileForContext=False
if AddAFileForContext:
collength, colupload = st.columns([2,3]) # adjust the ratio as needed
with collength:
#max_length = st.slider(key='maxlength', label="File section length for large files", min_value=1000, max_value=128000, value=12000, step=1000)
max_length = 128000
with colupload:
uploaded_file = st.file_uploader("Add a file for context:", type=["pdf", "xml", "json", "xlsx", "csv", "html", "htm", "md", "txt"])
document_sections = deque()
document_responses = {}
if uploaded_file is not None:
file_content = read_file_content(uploaded_file, max_length)
document_sections.extend(divide_document(file_content, max_length))
if len(document_sections) > 0:
if st.button("👁️ View Upload"):
st.markdown("**Sections of the uploaded file:**")
for i, section in enumerate(list(document_sections)):
st.markdown(f"**Section {i+1}**\n{section}")
st.markdown("**Chat with the model:**")
for i, section in enumerate(list(document_sections)):
if i in document_responses:
st.markdown(f"**Section {i+1}**\n{document_responses[i]}")
else:
if st.button(f"Chat about Section {i+1}"):
st.write('Reasoning with your inputs...')
st.write('Response:')
st.write(response)
document_responses[i] = response
filename = generate_filename(f"{user_prompt}_section_{i+1}", choice)
create_file(filename, user_prompt, response, should_save)
st.sidebar.markdown(get_table_download_link(filename), unsafe_allow_html=True)
num_columns_video=st.slider(key="num_columns_video", label="Choose Number of Video Columns", min_value=1, max_value=15, value=4)
display_videos_and_links(num_columns_video) # Video Jump Grid
num_columns_images=st.slider(key="num_columns_images", label="Choose Number of Image Columns", min_value=1, max_value=15, value=4)
display_images_and_wikipedia_summaries(num_columns_images) # Image Jump Grid
display_glossary_grid(roleplaying_glossary) # Word Glossary Jump Grid - Dynamically calculates columns based on details length to keep topic together
num_columns_text=st.slider(key="num_columns_text", label="Choose Number of Text Columns", min_value=1, max_value=15, value=4)
display_buttons_with_scores(num_columns_text) # Feedback Jump Grid
|