Spaces:
Running
Running
File size: 110,507 Bytes
01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 fc2b4e3 01531b8 |
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 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 |
"""
Trackio Deployment on Hugging Face Spaces
A Gradio interface for experiment tracking and monitoring
"""
import gradio as gr
import os
import json
import logging
from datetime import datetime
from typing import Dict, Any, Optional
import requests
import plotly.graph_objects as go
import plotly.express as px
import pandas as pd
import numpy as np
import plotly.io as pio
pio.templates.default = "plotly_white"
# Setup logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
class TrackioSpace:
"""Trackio deployment for Hugging Face Spaces using HF Datasets"""
def __init__(self, hf_token: Optional[str] = None, dataset_repo: Optional[str] = None):
self.experiments = {}
self.current_experiment = None
self.using_backup_data = False
# Get dataset repository and HF token from parameters or environment variables
self.dataset_repo = dataset_repo or os.environ.get('TRACKIO_DATASET_REPO', 'Tonic/trackio-experiments')
self.hf_token = hf_token or os.environ.get('HF_TOKEN')
# Initialize dataset manager for safe operations
self.dataset_manager = None
if self.hf_token and self.dataset_repo:
try:
# Prefer local dataset_utils in Space repo
from dataset_utils import TrackioDatasetManager # type: ignore
self.dataset_manager = TrackioDatasetManager(self.dataset_repo, self.hf_token)
logger.info("✅ Dataset manager initialized for safe operations (local)")
except Exception as local_e:
try:
# Fallback: try project src layout if present
import sys
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', '..', 'src'))
from dataset_utils import TrackioDatasetManager # type: ignore
self.dataset_manager = TrackioDatasetManager(self.dataset_repo, self.hf_token)
logger.info("✅ Dataset manager initialized for safe operations (fallback src)")
except Exception as e:
logger.warning(f"⚠️ Dataset manager not available, using legacy data handling: {local_e or e}")
logger.info(f"🔧 Using dataset repository: {self.dataset_repo}")
if not self.hf_token:
logger.warning("⚠️ HF_TOKEN not found. Some features may not work.")
self._load_experiments()
def _load_experiments(self):
"""Load experiments from HF Dataset with data preservation support"""
try:
# Try using dataset manager first for safe operations
if self.dataset_manager:
logger.info("🔍 Loading experiments using dataset manager")
experiments_list = self.dataset_manager.load_existing_experiments()
# Convert list to dict format expected by the interface
self.experiments = {}
for exp_data in experiments_list:
exp_id = exp_data.get('experiment_id')
if exp_id:
converted_experiment = self._convert_dataset_row_to_experiment(exp_data)
if converted_experiment:
self.experiments[exp_id] = converted_experiment
logger.info(f"✅ Loaded {len(self.experiments)} experiments using dataset manager")
# Sort experiments by creation date (newest first)
self.experiments = dict(sorted(
self.experiments.items(),
key=lambda x: x[1].get('created_at', ''),
reverse=True
))
# If no experiments found, use backup but mark backup mode to avoid accidental writes
if not self.experiments:
logger.info("📊 No experiments found in dataset, using backup data")
self._load_backup_experiments()
self.using_backup_data = True
return
# Fallback to direct dataset loading if dataset manager not available
if self.hf_token:
success = self._load_experiments_direct()
if success:
self.using_backup_data = False
return
# Final fallback to backup data
logger.info("🔄 Using backup data")
self._load_backup_experiments()
self.using_backup_data = True
except Exception as e:
logger.error(f"❌ Failed to load experiments: {e}")
self._load_backup_experiments()
self.using_backup_data = True
def _load_experiments_direct(self) -> bool:
"""Load experiments directly from HF Dataset without dataset manager"""
try:
from datasets import load_dataset
logger.info(f"🔍 Loading experiments directly from {self.dataset_repo}")
try:
dataset = load_dataset(self.dataset_repo, token=self.hf_token)
except Exception:
# Relax verification to handle split metadata mismatches
dataset = load_dataset(self.dataset_repo, token=self.hf_token, verification_mode="no_checks") # type: ignore[arg-type]
logger.info(f"✅ Successfully loaded dataset from {self.dataset_repo}")
# Convert dataset to experiments dict
self.experiments = {}
if 'train' in dataset:
for row in dataset['train']:
exp_id = row.get('experiment_id')
if exp_id:
converted_experiment = self._convert_dataset_row_to_experiment(row)
if converted_experiment:
self.experiments[exp_id] = converted_experiment
logger.info(f"📊 Successfully loaded {len(self.experiments)} experiments from dataset")
# Sort experiments by creation date (newest first)
self.experiments = dict(sorted(
self.experiments.items(),
key=lambda x: x[1].get('created_at', ''),
reverse=True
))
return True
except Exception as e:
logger.warning(f"⚠️ Failed to load from dataset directly: {e}")
return False
def _convert_dataset_row_to_experiment(self, row: Dict[str, Any]) -> Optional[Dict[str, Any]]:
"""Convert a dataset row to experiment format, handling JSON parsing safely"""
try:
exp_id = row.get('experiment_id')
if not exp_id:
return None
# Parse JSON fields safely
try:
metrics_raw = row.get('metrics', '[]')
if isinstance(metrics_raw, str):
metrics = json.loads(metrics_raw) if metrics_raw else []
else:
metrics = metrics_raw if metrics_raw else []
parameters_raw = row.get('parameters', '{}')
if isinstance(parameters_raw, str):
parameters = json.loads(parameters_raw) if parameters_raw else {}
else:
parameters = parameters_raw if parameters_raw else {}
artifacts_raw = row.get('artifacts', '[]')
if isinstance(artifacts_raw, str):
artifacts = json.loads(artifacts_raw) if artifacts_raw else []
else:
artifacts = artifacts_raw if artifacts_raw else []
logs_raw = row.get('logs', '[]')
if isinstance(logs_raw, str):
logs = json.loads(logs_raw) if logs_raw else []
else:
logs = logs_raw if logs_raw else []
except json.JSONDecodeError as json_err:
logger.warning(f"JSON decode error for experiment {exp_id}: {json_err}")
metrics, parameters, artifacts, logs = [], {}, [], []
return {
'id': exp_id,
'name': row.get('name', ''),
'description': row.get('description', ''),
'created_at': row.get('created_at', ''),
'status': row.get('status', 'running'),
'metrics': metrics,
'parameters': parameters,
'artifacts': artifacts,
'logs': logs,
'last_updated': row.get('last_updated', '')
}
except Exception as e:
logger.warning(f"Failed to convert dataset row to experiment: {e}")
return None
def _load_backup_experiments(self):
"""Load backup experiments when dataset is not available"""
logger.info("🔄 Loading backup experiments...")
backup_experiments = {
'exp_20250720_130853': {
'id': 'exp_20250720_130853',
'name': 'petite-elle-l-aime-3',
'description': 'SmolLM3 fine-tuning experiment',
'created_at': '2025-07-20T11:20:01.780908',
'status': 'running',
'metrics': [
{
'timestamp': '2025-07-20T11:20:01.780908',
'step': 25,
'metrics': {
'loss': 1.1659,
'grad_norm': 10.3125,
'learning_rate': 7e-08,
'num_tokens': 1642080.0,
'mean_token_accuracy': 0.75923578992486,
'epoch': 0.004851130919895701
}
},
{
'timestamp': '2025-07-20T11:26:39.042155',
'step': 50,
'metrics': {
'loss': 1.165,
'grad_norm': 10.75,
'learning_rate': 1.4291666666666667e-07,
'num_tokens': 3324682.0,
'mean_token_accuracy': 0.7577659255266189,
'epoch': 0.009702261839791402
}
},
{
'timestamp': '2025-07-20T11:33:16.203045',
'step': 75,
'metrics': {
'loss': 1.1639,
'grad_norm': 10.6875,
'learning_rate': 2.1583333333333334e-07,
'num_tokens': 4987941.0,
'mean_token_accuracy': 0.7581205774843692,
'epoch': 0.014553392759687101
}
},
{
'timestamp': '2025-07-20T11:39:53.453917',
'step': 100,
'metrics': {
'loss': 1.1528,
'grad_norm': 10.75,
'learning_rate': 2.8875e-07,
'num_tokens': 6630190.0,
'mean_token_accuracy': 0.7614579878747463,
'epoch': 0.019404523679582803
}
}
],
'parameters': {
'model_name': 'HuggingFaceTB/SmolLM3-3B',
'max_seq_length': 12288,
'use_flash_attention': True,
'use_gradient_checkpointing': False,
'batch_size': 8,
'gradient_accumulation_steps': 16,
'learning_rate': 3.5e-06,
'weight_decay': 0.01,
'warmup_steps': 1200,
'max_iters': 18000,
'eval_interval': 1000,
'log_interval': 25,
'save_interval': 2000,
'optimizer': 'adamw_torch',
'beta1': 0.9,
'beta2': 0.999,
'eps': 1e-08,
'scheduler': 'cosine',
'min_lr': 3.5e-07,
'fp16': False,
'bf16': True,
'ddp_backend': 'nccl',
'ddp_find_unused_parameters': False,
'save_steps': 2000,
'eval_steps': 1000,
'logging_steps': 25,
'save_total_limit': 5,
'eval_strategy': 'steps',
'metric_for_best_model': 'eval_loss',
'greater_is_better': False,
'load_best_model_at_end': True,
'data_dir': None,
'train_file': None,
'validation_file': None,
'test_file': None,
'use_chat_template': True,
'chat_template_kwargs': {'add_generation_prompt': True, 'no_think_system_message': True},
'enable_tracking': True,
'trackio_url': 'https://tonic-test-trackio-test.hf.space',
'trackio_token': None,
'log_artifacts': True,
'log_metrics': True,
'log_config': True,
'experiment_name': 'petite-elle-l-aime-3',
'dataset_name': 'legmlai/openhermes-fr',
'dataset_split': 'train',
'input_field': 'prompt',
'target_field': 'accepted_completion',
'filter_bad_entries': True,
'bad_entry_field': 'bad_entry',
'packing': False,
'max_prompt_length': 12288,
'max_completion_length': 8192,
'truncation': True,
'dataloader_num_workers': 10,
'dataloader_pin_memory': True,
'dataloader_prefetch_factor': 3,
'max_grad_norm': 1.0,
'group_by_length': True
},
'artifacts': [],
'logs': []
},
'exp_20250720_134319': {
'id': 'exp_20250720_134319',
'name': 'petite-elle-l-aime-3-1',
'description': 'SmolLM3 fine-tuning experiment',
'created_at': '2025-07-20T11:54:31.993219',
'status': 'running',
'metrics': [
{
'timestamp': '2025-07-20T11:54:31.993219',
'step': 25,
'metrics': {
'loss': 1.166,
'grad_norm': 10.375,
'learning_rate': 7e-08,
'num_tokens': 1642080.0,
'mean_token_accuracy': 0.7590958896279335,
'epoch': 0.004851130919895701,
'gpu_0_memory_allocated': 17.202261447906494,
'gpu_0_memory_reserved': 75.474609375,
'gpu_0_utilization': 0,
'cpu_percent': 2.7,
'memory_percent': 10.1
}
},
{
'timestamp': '2025-07-20T11:54:33.589487',
'step': 25,
'metrics': {
'gpu_0_memory_allocated': 17.202261447906494,
'gpu_0_memory_reserved': 75.474609375,
'gpu_0_utilization': 0,
'cpu_percent': 2.7,
'memory_percent': 10.1
}
}
],
'parameters': {
'model_name': 'HuggingFaceTB/SmolLM3-3B',
'max_seq_length': 12288,
'use_flash_attention': True,
'use_gradient_checkpointing': False,
'batch_size': 8,
'gradient_accumulation_steps': 16,
'learning_rate': 3.5e-06,
'weight_decay': 0.01,
'warmup_steps': 1200,
'max_iters': 18000,
'eval_interval': 1000,
'log_interval': 25,
'save_interval': 2000,
'optimizer': 'adamw_torch',
'beta1': 0.9,
'beta2': 0.999,
'eps': 1e-08,
'scheduler': 'cosine',
'min_lr': 3.5e-07,
'fp16': False,
'bf16': True,
'ddp_backend': 'nccl',
'ddp_find_unused_parameters': False,
'save_steps': 2000,
'eval_steps': 1000,
'logging_steps': 25,
'save_total_limit': 5,
'eval_strategy': 'steps',
'metric_for_best_model': 'eval_loss',
'greater_is_better': False,
'load_best_model_at_end': True,
'data_dir': None,
'train_file': None,
'validation_file': None,
'test_file': None,
'use_chat_template': True,
'chat_template_kwargs': {'add_generation_prompt': True, 'no_think_system_message': True},
'enable_tracking': True,
'trackio_url': 'https://tonic-test-trackio-test.hf.space',
'trackio_token': None,
'log_artifacts': True,
'log_metrics': True,
'log_config': True,
'experiment_name': 'petite-elle-l-aime-3-1',
'dataset_name': 'legmlai/openhermes-fr',
'dataset_split': 'train',
'input_field': 'prompt',
'target_field': 'accepted_completion',
'filter_bad_entries': True,
'bad_entry_field': 'bad_entry',
'packing': False,
'max_prompt_length': 12288,
'max_completion_length': 8192,
'truncation': True,
'dataloader_num_workers': 10,
'dataloader_pin_memory': True,
'dataloader_prefetch_factor': 3,
'max_grad_norm': 1.0,
'group_by_length': True
},
'artifacts': [],
'logs': []
}
}
self.experiments = backup_experiments
self.current_experiment = 'exp_20250720_134319'
logger.info(f"✅ Loaded {len(backup_experiments)} backup experiments")
def _save_experiments(self):
"""Save experiments to HF Dataset with data preservation
Note: This saves the full in-memory set. Prefer per-operation upsert via
dataset manager when available to reduce overwrite risk.
"""
try:
if self.using_backup_data:
logger.warning("⚠️ Using backup data; skip saving to dataset to avoid overwriting with demo values")
return
# Use dataset manager for safe operations if available
if self.dataset_manager:
logger.info("💾 Saving experiments using dataset manager (data preservation)")
# Convert current experiments to dataset format
experiments_to_save = []
for exp_id, exp_data in self.experiments.items():
experiment_entry = {
'experiment_id': exp_id,
'name': exp_data.get('name', ''),
'description': exp_data.get('description', ''),
'created_at': exp_data.get('created_at', ''),
'status': exp_data.get('status', 'running'),
'metrics': json.dumps(exp_data.get('metrics', []), default=str),
'parameters': json.dumps(exp_data.get('parameters', {}), default=str),
'artifacts': json.dumps(exp_data.get('artifacts', []), default=str),
'logs': json.dumps(exp_data.get('logs', []), default=str),
'last_updated': datetime.now().isoformat()
}
experiments_to_save.append(experiment_entry)
# Use dataset manager to save with data preservation
success = self.dataset_manager.save_experiments(
experiments_to_save,
f"Update experiments from Trackio Space ({len(experiments_to_save)} total experiments)"
)
if success:
logger.info(f"✅ Successfully saved {len(experiments_to_save)} experiments with data preservation")
else:
logger.error("❌ Failed to save experiments using dataset manager")
# Fallback to legacy method
self._save_experiments_legacy()
return
# Fallback to legacy method if dataset manager not available
self._save_experiments_legacy()
except Exception as e:
logger.error(f"❌ Failed to save experiments: {e}")
# Fallback to legacy method
self._save_experiments_legacy()
def _upsert_experiment(self, experiment_id: str):
"""Non-destructive upsert of a single experiment when dataset manager is available."""
try:
if not self.dataset_manager:
# Fallback to legacy save of full set
self._save_experiments()
return
exp = self.experiments.get(experiment_id)
if not exp:
return
payload = {
'experiment_id': experiment_id,
'name': exp.get('name', ''),
'description': exp.get('description', ''),
'created_at': exp.get('created_at', ''),
'status': exp.get('status', 'running'),
'metrics': json.dumps(exp.get('metrics', []), default=str),
'parameters': json.dumps(exp.get('parameters', {}), default=str),
'artifacts': json.dumps(exp.get('artifacts', []), default=str),
'logs': json.dumps(exp.get('logs', []), default=str),
'last_updated': datetime.now().isoformat()
}
self.dataset_manager.upsert_experiment(payload)
except Exception as e:
logger.warning(f"⚠️ Upsert failed, falling back to legacy save: {e}")
self._save_experiments()
def _save_experiments_legacy(self):
"""Legacy save method without data preservation (fallback only)"""
try:
if self.hf_token:
from datasets import Dataset
from huggingface_hub import HfApi
logger.warning("⚠️ Using legacy save method - data preservation not guaranteed")
# Convert experiments to dataset format
dataset_data = []
for exp_id, exp_data in self.experiments.items():
dataset_data.append({
'experiment_id': exp_id,
'name': exp_data.get('name', ''),
'description': exp_data.get('description', ''),
'created_at': exp_data.get('created_at', ''),
'status': exp_data.get('status', 'running'),
'metrics': json.dumps(exp_data.get('metrics', []), default=str),
'parameters': json.dumps(exp_data.get('parameters', {}), default=str),
'artifacts': json.dumps(exp_data.get('artifacts', []), default=str),
'logs': json.dumps(exp_data.get('logs', []), default=str),
'last_updated': datetime.now().isoformat()
})
# Create dataset
dataset = Dataset.from_list(dataset_data)
# Push to HF Hub
api = HfApi(token=self.hf_token)
dataset.push_to_hub(
self.dataset_repo,
token=self.hf_token,
private=True,
commit_message=f"Legacy update: {len(dataset_data)} experiments"
)
logger.info(f"✅ Saved {len(dataset_data)} experiments to {self.dataset_repo} (legacy method)")
else:
logger.warning("⚠️ No HF_TOKEN available, experiments not saved to dataset")
except Exception as e:
logger.error(f"❌ Failed to save experiments with legacy method: {e}")
# Fall back to local file for backup
try:
data = {
'experiments': self.experiments,
'current_experiment': self.current_experiment,
'last_updated': datetime.now().isoformat()
}
with open("trackio_experiments_backup.json", 'w') as f:
json.dump(data, f, indent=2, default=str)
logger.info("✅ Saved backup to local file")
except Exception as backup_e:
logger.error(f"❌ Failed to save backup: {backup_e}")
def create_experiment(self, name: str, description: str = "") -> Dict[str, Any]:
"""Create a new experiment"""
experiment_id = f"exp_{datetime.now().strftime('%Y%m%d_%H%M%S')}"
experiment = {
'id': experiment_id,
'name': name,
'description': description,
'created_at': datetime.now().isoformat(),
'status': 'running',
'metrics': [],
'parameters': {},
'artifacts': [],
'logs': []
}
self.experiments[experiment_id] = experiment
self.current_experiment = experiment_id
self._upsert_experiment(experiment_id)
logger.info(f"Created experiment: {experiment_id} - {name}")
return experiment
def log_metrics(self, experiment_id: str, metrics: Dict[str, Any], step: Optional[int] = None):
"""Log metrics for an experiment"""
if experiment_id not in self.experiments:
raise ValueError(f"Experiment {experiment_id} not found")
metric_entry = {
'timestamp': datetime.now().isoformat(),
'step': step,
'metrics': metrics
}
self.experiments[experiment_id]['metrics'].append(metric_entry)
self._upsert_experiment(experiment_id)
logger.info(f"Logged metrics for experiment {experiment_id}: {metrics}")
def log_parameters(self, experiment_id: str, parameters: Dict[str, Any]):
"""Log parameters for an experiment"""
if experiment_id not in self.experiments:
raise ValueError(f"Experiment {experiment_id} not found")
self.experiments[experiment_id]['parameters'].update(parameters)
self._upsert_experiment(experiment_id)
logger.info(f"Logged parameters for experiment {experiment_id}: {parameters}")
def log_artifact(self, experiment_id: str, artifact_name: str, artifact_data: str):
"""Log an artifact for an experiment"""
if experiment_id not in self.experiments:
raise ValueError(f"Experiment {experiment_id} not found")
artifact_entry = {
'name': artifact_name,
'timestamp': datetime.now().isoformat(),
'data': artifact_data
}
self.experiments[experiment_id]['artifacts'].append(artifact_entry)
self._save_experiments()
logger.info(f"Logged artifact for experiment {experiment_id}: {artifact_name}")
def get_experiment(self, experiment_id: str) -> Optional[Dict[str, Any]]:
"""Get experiment details"""
return self.experiments.get(experiment_id)
def list_experiments(self) -> Dict[str, Any]:
"""List all experiments"""
return {
'experiments': list(self.experiments.keys()),
'current_experiment': self.current_experiment,
'total_experiments': len(self.experiments)
}
def update_experiment_status(self, experiment_id: str, status: str):
"""Update experiment status"""
if experiment_id in self.experiments:
self.experiments[experiment_id]['status'] = status
self._upsert_experiment(experiment_id)
logger.info(f"Updated experiment {experiment_id} status to {status}")
def get_metrics_dataframe(self, experiment_id: str) -> pd.DataFrame:
"""Get metrics as a pandas DataFrame for plotting"""
if experiment_id not in self.experiments:
return pd.DataFrame()
experiment = self.experiments[experiment_id]
if not experiment['metrics']:
return pd.DataFrame()
# Convert metrics to DataFrame (merge duplicate steps)
data = []
for metric_entry in experiment['metrics']:
step = metric_entry.get('step', 0)
timestamp = metric_entry.get('timestamp', '')
metrics = metric_entry.get('metrics', {})
row = {'step': step, 'timestamp': timestamp}
row.update(metrics)
data.append(row)
if not data:
return pd.DataFrame()
df = pd.DataFrame(data)
# Ensure step exists even if None
if 'step' not in df.columns:
df['step'] = 0
# For duplicate steps, keep the latest timestamp and merge columns by last valid value
try:
df.sort_values(['step', 'timestamp'], inplace=True)
# Take the last row per step (latest timestamp)
df = df.groupby('step', as_index=False).last()
except Exception:
pass
return df
# Global instance
trackio_space = TrackioSpace()
def update_trackio_config(hf_token: str, dataset_repo: str) -> str:
"""Update TrackioSpace configuration with new HF token and dataset repository"""
global trackio_space
try:
# Create new instance with updated configuration
trackio_space = TrackioSpace(hf_token=hf_token if hf_token.strip() else None,
dataset_repo=dataset_repo if dataset_repo.strip() else None)
# Reload experiments with new configuration
trackio_space._load_experiments()
# Check if dataset manager is available
manager_status = "✅ Available (data preservation enabled)" if trackio_space.dataset_manager else "⚠️ Not available (legacy mode)"
return f"✅ Configuration updated successfully!\n📊 Dataset: {trackio_space.dataset_repo}\n🔑 HF Token: {'Set' if trackio_space.hf_token else 'Not set'}\n🛡️ Data Manager: {manager_status}\n📈 Loaded {len(trackio_space.experiments)} experiments"
except Exception as e:
return f"❌ Failed to update configuration: {str(e)}"
def test_dataset_connection(hf_token: str, dataset_repo: str) -> str:
"""Test connection to HF Dataset repository"""
try:
if not hf_token.strip():
return "❌ Please provide a Hugging Face token"
if not dataset_repo.strip():
return "❌ Please provide a dataset repository"
from datasets import load_dataset
# Test loading the dataset (with relaxed verification fallback)
try:
dataset = load_dataset(dataset_repo, token=hf_token)
except Exception:
dataset = load_dataset(dataset_repo, token=hf_token, verification_mode="no_checks") # type: ignore[arg-type]
# Count experiments and analyze structure
experiment_count = len(dataset['train']) if 'train' in dataset else 0
# Get column information
columns = list(dataset['train'].column_names) if 'train' in dataset else []
# Sample first few experiment IDs
sample_experiments = []
if 'train' in dataset and experiment_count > 0:
for i, row in enumerate(dataset['train']):
if i >= 3: # Only show first 3
break
sample_experiments.append(row.get('experiment_id', 'unknown'))
result = f"✅ Connection successful!\n📊 Dataset: {dataset_repo}\n📈 Found {experiment_count} experiments\n🔗 Dataset URL: https://huggingface.co/datasets/{dataset_repo}\n\n"
result += f"📋 Dataset Columns: {', '.join(columns)}\n"
if sample_experiments:
result += f"🔬 Sample Experiments: {', '.join(sample_experiments)}\n"
# Test parsing one experiment if available
if 'train' in dataset and experiment_count > 0:
first_row = dataset['train'][0]
exp_id = first_row.get('experiment_id', 'unknown')
metrics_raw = first_row.get('metrics', '[]')
try:
if isinstance(metrics_raw, str):
metrics = json.loads(metrics_raw)
metrics_count = len(metrics) if isinstance(metrics, list) else 0
result += f"📊 First experiment ({exp_id}) metrics: {metrics_count} entries\n"
else:
result += f"📊 First experiment ({exp_id}) metrics: Non-string format\n"
except json.JSONDecodeError as e:
result += f"⚠️ JSON parse error in first experiment: {e}\n"
return result
except Exception as e:
return f"❌ Connection failed: {str(e)}\n\n💡 Troubleshooting:\n1. Check your HF token is correct\n2. Verify the dataset repository exists\n3. Ensure your token has read access to the dataset"
def create_dataset_repository(hf_token: str, dataset_repo: str) -> str:
"""Create HF Dataset repository if it doesn't exist"""
try:
if not hf_token.strip():
return "❌ Please provide a Hugging Face token"
if not dataset_repo.strip():
return "❌ Please provide a dataset repository"
from datasets import Dataset
from huggingface_hub import HfApi
# Parse username and dataset name
if '/' not in dataset_repo:
return "❌ Dataset repository must be in format: username/dataset-name"
username, dataset_name = dataset_repo.split('/', 1)
# Create API client
api = HfApi(token=hf_token)
# Check if dataset exists
try:
api.dataset_info(dataset_repo)
return f"✅ Dataset {dataset_repo} already exists!\n🛡️ Data preservation is enabled for existing datasets\n🔗 View at: https://huggingface.co/datasets/{dataset_repo}"
except:
# Dataset doesn't exist, create it
pass
# Try to initialize dataset manager to use its repository creation
try:
# Import dataset manager
import sys
sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', '..', 'src'))
from dataset_utils import TrackioDatasetManager
# Create dataset manager instance
dataset_manager = TrackioDatasetManager(dataset_repo, hf_token)
# Check if dataset exists using the manager
exists = dataset_manager.check_dataset_exists()
if exists:
return f"✅ Dataset {dataset_repo} already exists!\n🛡️ Data preservation is enabled\n🔗 View at: https://huggingface.co/datasets/{dataset_repo}"
except ImportError:
# Dataset manager not available, use legacy method
pass
except Exception as e:
# Dataset manager failed, use legacy method
logger.warning(f"Dataset manager failed: {e}, using legacy method")
# Create empty dataset with proper structure
empty_dataset = Dataset.from_dict({
'experiment_id': [],
'name': [],
'description': [],
'created_at': [],
'status': [],
'metrics': [],
'parameters': [],
'artifacts': [],
'logs': [],
'last_updated': []
})
# Push to hub
empty_dataset.push_to_hub(
dataset_repo,
token=hf_token,
private=True,
commit_message="Create Trackio experiment dataset with data preservation support"
)
return f"✅ Dataset {dataset_repo} created successfully!\n🛡️ Data preservation is now enabled\n🔗 View at: https://huggingface.co/datasets/{dataset_repo}\n📊 Ready to store experiments safely"
except Exception as e:
return f"❌ Failed to create dataset: {str(e)}\n\n💡 Troubleshooting:\n1. Check your HF token has write permissions\n2. Verify the username in the repository name\n3. Ensure the dataset name is valid\n4. Check internet connectivity"
"""
Initialize API client for remote data. We do not hardcode a default test URL to avoid
overwriting dataset content with demo data. The API client will only be initialized
when TRACKIO_URL or TRACKIO_SPACE_ID is present.
"""
api_client = None
try:
from trackio_api_client import TrackioAPIClient
# Resolve Trackio space from environment
trackio_url_env = os.environ.get('TRACKIO_URL') or os.environ.get('TRACKIO_SPACE_ID')
if trackio_url_env:
# Clean up URL to avoid double protocol issues
trackio_url = trackio_url_env
if trackio_url.startswith('https://https://'):
trackio_url = trackio_url.replace('https://https://', 'https://')
elif trackio_url.startswith('http://http://'):
trackio_url = trackio_url.replace('http://http://', 'http://')
api_client = TrackioAPIClient(trackio_url)
logger.info(f"✅ API client initialized for remote data access: {trackio_url}")
else:
logger.info("No TRACKIO_URL/TRACKIO_SPACE_ID set; remote API client disabled")
except ImportError:
logger.warning("⚠️ API client not available, using local data only")
except Exception as e:
logger.warning(f"⚠️ Failed to initialize API client: {e}, using local data only")
# Add Hugging Face Spaces compatibility
def is_huggingface_spaces():
"""Check if running on Hugging Face Spaces"""
return os.environ.get('SPACE_ID') is not None
def get_persistent_data_path():
"""Get a persistent data path for Hugging Face Spaces"""
if is_huggingface_spaces():
# Use a path that might persist better on HF Spaces
return "/tmp/trackio_experiments.json"
else:
return "trackio_experiments.json"
# Override the data file path for HF Spaces
if is_huggingface_spaces():
logger.info("🚀 Running on Hugging Face Spaces - using persistent storage")
trackio_space.data_file = get_persistent_data_path()
def get_remote_experiment_data(experiment_id: str) -> Dict[str, Any]:
"""Get experiment data from remote API"""
if api_client is None:
return None
try:
# Get experiment details from API
details_result = api_client.get_experiment_details(experiment_id)
if "success" in details_result:
return {"remote": True, "data": details_result["data"]}
else:
logger.warning(f"Failed to get remote data for {experiment_id}: {details_result}")
return None
except Exception as e:
logger.error(f"Error getting remote data: {e}")
return None
def parse_remote_metrics_data(experiment_details: str) -> pd.DataFrame:
"""Parse metrics data from remote experiment details"""
try:
# Look for metrics in the experiment details
lines = experiment_details.split('\n')
metrics_data = []
# First try to parse the new format with structured experiment details
for line in lines:
if 'Step:' in line and 'Metrics:' in line:
# Extract step and metrics from the line
try:
# Parse step number
step_part = line.split('Step:')[1].split('Metrics:')[0].strip()
step = int(step_part)
# Parse metrics JSON
metrics_part = line.split('Metrics:')[1].strip()
metrics = json.loads(metrics_part)
# Add timestamp
row = {'step': step, 'timestamp': datetime.now().isoformat()}
row.update(metrics)
metrics_data.append(row)
except (ValueError, json.JSONDecodeError) as e:
logger.warning(f"Failed to parse metrics line: {line} - {e}")
continue
# If no metrics found in text format, try to parse from the dataset directly
if not metrics_data:
logger.info("No metrics found in text format, trying to parse from experiment structure")
# This will be handled by the updated get_remote_experiment_data function
if metrics_data:
return pd.DataFrame(metrics_data)
else:
return pd.DataFrame()
except Exception as e:
logger.error(f"Error parsing remote metrics: {e}")
return pd.DataFrame()
def get_metrics_dataframe(experiment_id: str) -> pd.DataFrame:
"""Get metrics as a pandas DataFrame for plotting - tries dataset first, then local backup"""
try:
# First try to get data directly from the dataset using the dataset manager
if trackio_space.dataset_manager:
logger.info(f"Getting metrics for {experiment_id} from dataset")
experiment_data = trackio_space.dataset_manager.get_experiment_by_id(experiment_id)
if experiment_data:
# Parse metrics from the dataset
metrics_json = experiment_data.get('metrics', '[]')
if isinstance(metrics_json, str):
try:
metrics_list = json.loads(metrics_json)
# Convert to DataFrame format
df_data = []
for metric_entry in metrics_list:
if isinstance(metric_entry, dict):
step = metric_entry.get('step', 0)
timestamp = metric_entry.get('timestamp', '')
metrics = metric_entry.get('metrics', {})
row = {'step': step, 'timestamp': timestamp}
row.update(metrics)
df_data.append(row)
if df_data:
logger.info(f"Found {len(df_data)} metrics entries from dataset for {experiment_id}")
return pd.DataFrame(df_data)
else:
logger.warning(f"No valid metrics found in dataset for {experiment_id}")
except json.JSONDecodeError as e:
logger.warning(f"Failed to parse metrics JSON for {experiment_id}: {e}")
else:
logger.warning(f"Metrics data is not a JSON string for {experiment_id}")
else:
logger.warning(f"Experiment {experiment_id} not found in dataset")
# Try legacy remote data approach
remote_data = get_remote_experiment_data(experiment_id)
if remote_data:
logger.info(f"Using remote API data for {experiment_id}")
# Parse the remote experiment details to extract metrics
df = parse_remote_metrics_data(remote_data["data"])
if not df.empty:
logger.info(f"Found {len(df)} metrics entries from remote API")
return df
else:
logger.warning(f"No metrics found in remote API data for {experiment_id}")
# Fall back to local data
logger.info(f"Using local backup data for {experiment_id}")
return trackio_space.get_metrics_dataframe(experiment_id)
except Exception as e:
logger.error(f"Error getting metrics dataframe for {experiment_id}: {e}")
# Fall back to local data
logger.info(f"Falling back to local data for {experiment_id}")
return trackio_space.get_metrics_dataframe(experiment_id)
def create_experiment_interface(name: str, description: str):
"""Create a new experiment"""
try:
experiment = trackio_space.create_experiment(name, description)
# Return both the status message and a refreshed dropdown
msg = f"✅ Experiment created successfully!\nID: {experiment['id']}\nName: {experiment['name']}\nStatus: {experiment['status']}"
dropdown = gr.Dropdown(choices=get_experiment_dropdown_choices(), value=experiment['id'])
return msg, dropdown
except Exception as e:
dropdown = gr.Dropdown(choices=get_experiment_dropdown_choices(), value=None)
return f"❌ Error creating experiment: {str(e)}", dropdown
def log_metrics_interface(experiment_id: str, metrics_json: str, step: str) -> str:
"""Log metrics for an experiment"""
try:
metrics = json.loads(metrics_json)
step_int = int(step) if step else None
trackio_space.log_metrics(experiment_id, metrics, step_int)
return f"✅ Metrics logged successfully for experiment {experiment_id}\nStep: {step_int}\nMetrics: {json.dumps(metrics, indent=2)}"
except Exception as e:
return f"❌ Error logging metrics: {str(e)}"
def log_parameters_interface(experiment_id: str, parameters_json: str) -> str:
"""Log parameters for an experiment"""
try:
parameters = json.loads(parameters_json)
trackio_space.log_parameters(experiment_id, parameters)
return f"✅ Parameters logged successfully for experiment {experiment_id}\nParameters: {json.dumps(parameters, indent=2)}"
except Exception as e:
return f"❌ Error logging parameters: {str(e)}"
def get_experiment_details(experiment_id: str) -> str:
"""Get experiment details"""
try:
experiment = trackio_space.get_experiment(experiment_id)
if experiment:
# Format the output nicely
details = f"""
📊 EXPERIMENT DETAILS
====================
ID: {experiment['id']}
Name: {experiment['name']}
Description: {experiment['description']}
Status: {experiment['status']}
Created: {experiment['created_at']}
📈 METRICS COUNT: {len(experiment['metrics'])}
📋 PARAMETERS COUNT: {len(experiment['parameters'])}
📦 ARTIFACTS COUNT: {len(experiment['artifacts'])}
🔧 PARAMETERS:
{json.dumps(experiment['parameters'], indent=2)}
📊 LATEST METRICS:
"""
if experiment['metrics']:
latest_metrics = experiment['metrics'][-1]
details += f"Step: {latest_metrics.get('step', 'N/A')}\n"
details += f"Timestamp: {latest_metrics.get('timestamp', 'N/A')}\n"
details += f"Metrics: {json.dumps(latest_metrics.get('metrics', {}), indent=2)}"
else:
details += "No metrics logged yet."
return details
else:
return f"❌ Experiment {experiment_id} not found"
except Exception as e:
return f"❌ Error getting experiment details: {str(e)}"
def list_experiments_interface() -> str:
"""List all experiments with details"""
try:
experiments_info = trackio_space.list_experiments()
experiments = trackio_space.experiments
if not experiments:
return "📭 No experiments found. Create one first!"
result = f"📋 EXPERIMENTS OVERVIEW\n{'='*50}\n"
result += f"Total Experiments: {len(experiments)}\n"
result += f"Current Experiment: {experiments_info['current_experiment']}\n\n"
for exp_id, exp_data in experiments.items():
status_emoji = {
'running': '🟢',
'completed': '✅',
'failed': '❌',
'paused': '⏸️'
}.get(exp_data['status'], '❓')
result += f"{status_emoji} {exp_id}\n"
result += f" Name: {exp_data['name']}\n"
result += f" Status: {exp_data['status']}\n"
result += f" Created: {exp_data['created_at']}\n"
result += f" Metrics: {len(exp_data['metrics'])} entries\n"
result += f" Parameters: {len(exp_data['parameters'])} entries\n"
result += f" Artifacts: {len(exp_data['artifacts'])} entries\n\n"
return result
except Exception as e:
return f"❌ Error listing experiments: {str(e)}"
def update_experiment_status_interface(experiment_id: str, status: str) -> str:
"""Update experiment status"""
try:
trackio_space.update_experiment_status(experiment_id, status)
return f"✅ Experiment {experiment_id} status updated to {status}"
except Exception as e:
return f"❌ Error updating experiment status: {str(e)}"
def create_metrics_plot(experiment_id: str, metric_name: str = "loss") -> go.Figure:
"""Create a plot for a specific metric (supports all logged metrics, including new ones)"""
try:
df = get_metrics_dataframe(experiment_id)
if df.empty:
# Return empty plot
fig = go.Figure()
fig.add_annotation(
text="No metrics data available",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False
)
return fig
if metric_name not in df.columns:
# Show available metrics
available_metrics = [col for col in df.columns if col not in ['step', 'timestamp']]
fig = go.Figure()
fig.add_annotation(
text=f"Available metrics: {', '.join(available_metrics)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False
)
return fig
# Helper predicates
def _is_eval_metric(name: str) -> bool:
return name.startswith('eval_') or name.startswith('eval/')
def _is_system_metric(name: str) -> bool:
import re
if name in ("cpu_percent", "memory_percent"):
return True
return re.match(r"^gpu_\d+_(memory_allocated|memory_reserved|utilization)$", name) is not None
# Ensure steps are numeric and monotonically increasing to avoid zig-zag lines
try:
df = df.copy()
# Choose x-axis: time for system metrics, step otherwise
use_time_axis = _is_system_metric(metric_name)
if use_time_axis:
# Convert timestamp to datetime for nicer axis rendering
df['time'] = pd.to_datetime(df.get('timestamp', ''), errors='coerce')
# Fallback order if timestamps are missing
if df['time'].isna().all():
df['time'] = range(1, len(df) + 1)
df.sort_values('time', inplace=True)
x_field = 'time'
else:
# If step looks constant or missing, try to derive it from a common field
if 'step' not in df or df['step'].nunique() <= 1:
for alt in ['train/global_step', 'global_step', 'train/step']:
if alt in df.columns and df[alt].notna().any():
df['step'] = pd.to_numeric(df[alt], errors='coerce')
break
# If still missing or constant, fallback to an inferred counter by order of arrival
if 'step' not in df.columns or df['step'].isna().all() or df['step'].nunique() <= 1:
df['step'] = range(1, len(df) + 1)
else:
df['step'] = pd.to_numeric(df.get('step', -1), errors='coerce').fillna(-1)
df.sort_values('step', inplace=True)
x_field = 'step'
except Exception:
x_field = 'step'
# Filter rows where the metric is present to ensure connected lines
try:
plot_df = df[[x_field, metric_name]].dropna(subset=[metric_name]).copy()
except Exception:
plot_df = df
fig = px.line(plot_df, x=x_field, y=metric_name, title=f'{metric_name} over time')
fig.update_layout(
xaxis_title="Time" if (metric_name in ("cpu_percent", "memory_percent") or metric_name.startswith('gpu_')) else "Training Step",
yaxis_title=metric_name.title(),
hovermode='x unified'
)
# Connect points for evaluation metrics, avoid connecting gaps for others
try:
for trace in fig.data:
trace.connectgaps = True if _is_eval_metric(metric_name) else False
# Force line+markers to visually connect points
trace.mode = 'lines+markers'
except Exception:
pass
return fig
except Exception as e:
fig = go.Figure()
fig.add_annotation(
text=f"Error creating plot: {str(e)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False
)
return fig
def create_experiment_comparison(experiment_ids: str) -> go.Figure:
"""Compare multiple experiments"""
try:
exp_ids = [exp_id.strip() for exp_id in experiment_ids.split(',')]
fig = go.Figure()
for exp_id in exp_ids:
df = get_metrics_dataframe(exp_id)
if not df.empty and 'loss' in df.columns:
fig.add_trace(go.Scatter(
x=df['step'],
y=df['loss'],
mode='lines+markers',
name=f"{exp_id} - Loss",
line=dict(width=2)
))
fig.update_layout(
title="Experiment Comparison - Loss",
xaxis_title="Training Step",
yaxis_title="Loss",
hovermode='x unified'
)
return fig
except Exception as e:
fig = go.Figure()
fig.add_annotation(
text=f"Error creating comparison: {str(e)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False
)
return fig
def simulate_training_data(experiment_id: str):
"""Simulate training data for demonstration"""
try:
import random
import time
last_time = time.time()
for step in range(0, 1000, 50):
# Simulate loss decreasing over time
loss = 2.0 * np.exp(-step / 500) + 0.1 * np.random.random()
accuracy = 0.3 + 0.6 * (1 - np.exp(-step / 300)) + 0.05 * np.random.random()
lr = 3.5e-6 * (0.9 ** (step // 200))
batch_size = 8
seq_len = 2048
total_tokens = batch_size * seq_len
padding_tokens = random.randint(0, batch_size * 32)
truncated_tokens = random.randint(0, batch_size * 8)
now = time.time()
step_time = random.uniform(0.4, 0.7)
throughput = total_tokens / step_time
token_acc = accuracy
gate_ortho = random.uniform(0.01, 0.05)
center = random.uniform(0.01, 0.05)
metrics = {
"loss": round(loss, 4),
"accuracy": round(accuracy, 4),
"learning_rate": round(lr, 8),
"gpu_memory": round(20 + 5 * np.random.random(), 2),
"training_time": round(0.5 + 0.2 * np.random.random(), 3),
"total_tokens": total_tokens,
"padding_tokens": padding_tokens,
"truncated_tokens": truncated_tokens,
"throughput": throughput,
"step_time": step_time,
"batch_size": batch_size,
"seq_len": seq_len,
"token_acc": token_acc,
"train/gate_ortho": gate_ortho,
"train/center": center
}
trackio_space.log_metrics(experiment_id, metrics, step)
last_time = now
return f"✅ Simulated training data for experiment {experiment_id}\nAdded 20 metric entries (steps 0-950)"
except Exception as e:
return f"❌ Error simulating data: {str(e)}"
def create_demo_experiment():
"""Create a demo experiment with training data"""
try:
# Create demo experiment
experiment = trackio_space.create_experiment(
"demo_smollm3_training",
"Demo experiment with simulated training data"
)
experiment_id = experiment['id']
# Add some demo parameters
parameters = {
"model_name": "HuggingFaceTB/SmolLM3-3B",
"batch_size": 8,
"learning_rate": 3.5e-6,
"max_iters": 18000,
"mixed_precision": "bf16",
"dataset": "legmlai/openhermes-fr"
}
trackio_space.log_parameters(experiment_id, parameters)
# Add demo training data
simulate_training_data(experiment_id)
return f"✅ Demo experiment created: {experiment_id}\nYou can now test the visualization with this experiment!"
except Exception as e:
return f"❌ Error creating demo experiment: {str(e)}"
# Helper functions for the new interface
def get_experiment_dropdown_choices() -> list:
"""Get the list of experiments for the dropdown"""
experiments = list(trackio_space.experiments.keys())
if not experiments:
return ["No experiments available"]
return experiments
def refresh_experiment_dropdown() -> tuple:
"""Refresh the experiment dropdown and return current choices"""
choices = get_experiment_dropdown_choices()
current_value = choices[0] if choices and choices[0] != "No experiments available" else None
return gr.Dropdown(choices=choices, value=current_value)
def get_available_metrics_for_experiments(experiment_ids: list) -> list:
"""Get all available metrics across selected experiments"""
try:
all_metrics = set()
for exp_id in experiment_ids:
df = get_metrics_dataframe(exp_id)
if not df.empty:
# Get numeric columns (excluding step and timestamp)
numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist()
numeric_cols = [col for col in numeric_cols if col not in ['step']]
all_metrics.update(numeric_cols)
return sorted(list(all_metrics))
except Exception as e:
logger.error(f"Error getting available metrics: {str(e)}")
return ["loss", "accuracy"]
def create_test_plot() -> go.Figure:
"""Create a simple test plot to verify plotly rendering works"""
try:
# Create simple test data
x = [1, 2, 3, 4, 5]
y = [1, 4, 2, 3, 5]
fig = go.Figure()
fig.add_trace(go.Scatter(
x=x,
y=y,
mode='lines+markers',
name='Test Data',
line=dict(width=2, color='blue'),
marker=dict(size=5, color='red'),
connectgaps=True,
hovertemplate='<b>X:</b> %{x}<br><b>Y:</b> %{y}<extra></extra>'
))
fig.update_layout(
title="Test Plot - If you can see this, plotly is working!",
xaxis_title="X Axis",
yaxis_title="Y Axis",
plot_bgcolor='white',
paper_bgcolor='white',
font=dict(size=14),
margin=dict(l=50, r=50, t=80, b=50)
)
fig.update_xaxes(showgrid=True, gridwidth=1, gridcolor='lightgray')
fig.update_yaxes(showgrid=True, gridwidth=1, gridcolor='lightgray')
logger.info("Test plot created successfully")
return fig
except Exception as e:
logger.error(f"Error creating test plot: {str(e)}")
fig = go.Figure()
fig.add_annotation(
text=f"Test plot error: {str(e)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=14, color="red")
)
return fig
def get_experiment_status_summary(experiment_id: str) -> str:
"""Get a formatted summary of experiment status and metadata"""
try:
experiment = trackio_space.get_experiment(experiment_id)
if not experiment:
return f"Experiment {experiment_id} not found."
summary = f"📋 EXPERIMENT STATUS SUMMARY\n{'='*50}\n"
summary += f"ID: {experiment['id']}\n"
summary += f"Name: {experiment['name']}\n"
summary += f"Description: {experiment['description']}\n"
summary += f"Status: {experiment['status']}\n"
summary += f"Created: {experiment['created_at']}\n"
summary += f"Metrics entries: {len(experiment['metrics'])}\n"
summary += f"Parameters: {len(experiment['parameters'])}\n"
summary += f"Artifacts: {len(experiment['artifacts'])}\n"
summary += f"Logs: {len(experiment['logs'])}\n"
# Add latest metrics if available
if experiment['metrics']:
latest = experiment['metrics'][-1]
summary += f"\n📈 LATEST METRICS (Step {latest.get('step', 'N/A')}):\n"
for k, v in latest.get('metrics', {}).items():
summary += f" {k}: {v}\n"
return summary
except Exception as e:
return f"Error generating status summary: {str(e)}"
def get_experiment_parameters_summary(experiment_id: str) -> str:
"""Get a formatted summary of experiment parameters"""
try:
experiment = trackio_space.get_experiment(experiment_id)
if not experiment:
return f"Experiment {experiment_id} not found."
params = experiment.get('parameters', {})
if not params:
return "No parameters logged for this experiment."
summary = f"🔧 PARAMETERS FOR {experiment_id}\n{'='*50}\n"
# Group parameters by category
model_params = {k: v for k, v in params.items() if 'model' in k.lower() or 'name' in k.lower()}
training_params = {k: v for k, v in params.items() if any(x in k.lower() for x in ['learning', 'batch', 'epoch', 'step', 'iter', 'optimizer'])}
data_params = {k: v for k, v in params.items() if any(x in k.lower() for x in ['data', 'dataset', 'file', 'split'])}
other_params = {k: v for k, v in params.items() if k not in model_params and k not in training_params and k not in data_params}
if model_params:
summary += "🤖 MODEL PARAMETERS:\n"
for k, v in model_params.items():
summary += f" {k}: {v}\n"
summary += "\n"
if training_params:
summary += "🏃 TRAINING PARAMETERS:\n"
for k, v in training_params.items():
summary += f" {k}: {v}\n"
summary += "\n"
if data_params:
summary += "📁 DATA PARAMETERS:\n"
for k, v in data_params.items():
summary += f" {k}: {v}\n"
summary += "\n"
if other_params:
summary += "⚙️ OTHER PARAMETERS:\n"
for k, v in other_params.items():
summary += f" {k}: {v}\n"
return summary
except Exception as e:
return f"Error generating parameters summary: {str(e)}"
def get_experiment_metrics_summary(experiment_id: str) -> str:
"""Get a summary of all metrics for an experiment"""
try:
df = get_metrics_dataframe(experiment_id)
if df.empty:
return "No metrics data available for this experiment.\n\n💡 This could mean:\n• The experiment hasn't started logging metrics yet\n• The experiment is using a different data format\n• No training has been performed on this experiment"
# Get numeric columns (excluding step and timestamp)
numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist()
numeric_cols = [col for col in numeric_cols if col not in ['step']]
if not numeric_cols:
return "No numeric metrics found for this experiment.\n\n💡 This could mean:\n• Only timestamp data is available\n• Metrics are stored in a different format\n• The experiment hasn't logged any numeric metrics yet"
summary = f"📊 METRICS SUMMARY FOR {experiment_id}\n{'='*50}\n"
summary += f"Total data points: {len(df)}\n"
summary += f"Steps range: {df['step'].min()} - {df['step'].max()}\n"
summary += f"Available metrics: {', '.join(numeric_cols)}\n\n"
for col in numeric_cols:
if col in df.columns:
values = df[col].dropna()
if len(values) > 0:
summary += f"{col}:\n"
summary += f" Min: {values.min():.6f}\n"
summary += f" Max: {values.max():.6f}\n"
summary += f" Mean: {values.mean():.6f}\n"
summary += f" Latest: {values.iloc[-1]:.6f}\n\n"
return summary
except Exception as e:
return f"Error generating metrics summary: {str(e)}"
def create_combined_metrics_plot(experiment_id: str) -> go.Figure:
"""Create a combined plot showing all metrics for an experiment"""
try:
if not experiment_id:
fig = go.Figure()
fig.add_annotation(
text="No experiment selected",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="gray")
)
fig.update_layout(
title="Select an Experiment",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
df = get_metrics_dataframe(experiment_id)
if df.empty:
fig = go.Figure()
fig.add_annotation(
text="No metrics data available for this experiment",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="red")
)
fig.update_layout(
title="No Data Available",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
# Get numeric columns (excluding step and timestamp)
numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist()
numeric_cols = [col for col in numeric_cols if col not in ['step']]
if not numeric_cols:
fig = go.Figure()
fig.add_annotation(
text="No numeric metrics found for this experiment",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="orange")
)
fig.update_layout(
title="No Metrics Found",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
# Create subplots for multiple metrics
from plotly.subplots import make_subplots
# Determine number of rows and columns for subplots
n_metrics = len(numeric_cols)
n_cols = min(3, n_metrics) # Max 3 columns
n_rows = (n_metrics + n_cols - 1) // n_cols
fig = make_subplots(
rows=n_rows, cols=n_cols,
subplot_titles=numeric_cols,
vertical_spacing=0.05,
horizontal_spacing=0.1
)
# Define colors for different metrics
colors = ['blue', 'red', 'green', 'orange', 'purple', 'brown', 'pink', 'gray', 'cyan', 'magenta']
# Helper predicates
def _is_eval_metric(name: str) -> bool:
return name.startswith('eval_') or name.startswith('eval/')
def _is_system_metric(name: str) -> bool:
import re
if name in ("cpu_percent", "memory_percent"):
return True
return re.match(r"^gpu_\d+_(memory_allocated|memory_reserved|utilization)$", name) is not None
for i, metric in enumerate(numeric_cols):
if metric in df.columns and not df[metric].isna().all():
row = (i // n_cols) + 1
col = (i % n_cols) + 1
color = colors[i % len(colors)]
# Clean steps for each subplot too
try:
df_sub = df.copy()
use_time_axis = _is_system_metric(metric)
if use_time_axis:
df_sub['time'] = pd.to_datetime(df_sub.get('timestamp', ''), errors='coerce')
if df_sub['time'].isna().all():
df_sub['time'] = range(1, len(df_sub) + 1)
df_sub.sort_values('time', inplace=True)
# Filter to available metric points only to ensure connected lines
metric_mask = df_sub[metric].notna()
x_vals = df_sub.loc[metric_mask, 'time'].tolist()
y_vals = df_sub.loc[metric_mask, metric].tolist()
else:
if 'step' not in df_sub or df_sub['step'].nunique() <= 1:
for alt in ['train/global_step', 'global_step', 'train/step']:
if alt in df_sub.columns and df_sub[alt].notna().any():
df_sub['step'] = pd.to_numeric(df_sub[alt], errors='coerce')
break
if 'step' not in df_sub.columns or df_sub['step'].isna().all() or df_sub['step'].nunique() <= 1:
df_sub['step'] = range(1, len(df_sub) + 1)
else:
df_sub['step'] = pd.to_numeric(df_sub.get('step', -1), errors='coerce').fillna(-1)
df_sub.sort_values('step', inplace=True)
# Filter to available metric points only to ensure connected lines
metric_mask = df_sub[metric].notna()
x_vals = df_sub.loc[metric_mask, 'step'].tolist()
y_vals = df_sub.loc[metric_mask, metric].tolist()
except Exception:
df_sub = df
metric_mask = df_sub[metric].notna() if metric in df_sub else []
x_vals = df_sub.get('step', list(range(1, len(df_sub) + 1))).tolist()
y_vals = df_sub.get(metric, []).tolist()
fig.add_trace(
go.Scatter(
x=x_vals,
y=y_vals,
mode='lines+markers',
name=metric,
line=dict(width=2, color=color),
marker=dict(size=4, color=color),
showlegend=False,
connectgaps=True if _is_eval_metric(metric) else False
),
row=row, col=col
)
# Set axis titles per subplot for clarity
try:
fig.update_xaxes(title_text=("Time" if use_time_axis else "Training Step"), row=row, col=col)
except Exception:
pass
fig.update_layout(
title=f"All Metrics for Experiment {experiment_id}",
height=350 * n_rows,
plot_bgcolor='white',
paper_bgcolor='white',
font=dict(size=12),
margin=dict(l=50, r=50, t=80, b=50)
)
# Update all subplot axes
for i in range(1, n_rows + 1):
for j in range(1, n_cols + 1):
fig.update_xaxes(
showgrid=True, gridwidth=1, gridcolor='lightgray',
zeroline=True, zerolinecolor='black',
row=i, col=j
)
fig.update_yaxes(
showgrid=True, gridwidth=1, gridcolor='lightgray',
zeroline=True, zerolinecolor='black',
row=i, col=j
)
return fig
except Exception as e:
logger.error(f"Error creating combined metrics plot: {str(e)}")
fig = go.Figure()
fig.add_annotation(
text=f"Error creating combined plot: {str(e)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=14, color="red")
)
return fig
def update_dashboard(experiment_id: str) -> tuple:
"""Update all dashboard components for a selected experiment"""
try:
if not experiment_id or experiment_id == "No experiments available":
return (
"Please select an experiment from the dropdown.",
"No experiment selected.",
"No experiment selected.",
create_combined_metrics_plot(""),
"No experiment selected."
)
# Get all the dashboard components
status_summary = get_experiment_status_summary(experiment_id)
parameters_summary = get_experiment_parameters_summary(experiment_id)
metrics_summary = get_experiment_metrics_summary(experiment_id)
combined_plot = create_combined_metrics_plot(experiment_id)
# Create a combined summary
combined_summary = f"{status_summary}\n\n{parameters_summary}\n\n{metrics_summary}"
return (
status_summary,
parameters_summary,
metrics_summary,
combined_plot,
combined_summary
)
except Exception as e:
error_msg = f"Error updating dashboard: {str(e)}"
return (error_msg, error_msg, error_msg, create_combined_metrics_plot(""), error_msg)
def update_dashboard_metric_plot(experiment_id: str, metric_name: str = "loss") -> go.Figure:
"""Update the dashboard metric plot for a selected experiment and metric"""
try:
if not experiment_id or experiment_id == "No experiments available":
return create_metrics_plot("", metric_name)
return create_metrics_plot(experiment_id, metric_name)
except Exception as e:
logger.error(f"Error updating dashboard metric plot: {str(e)}")
return create_metrics_plot("", metric_name)
def create_experiment_comparison_from_selection(selected_experiments: list, selected_metrics: list) -> go.Figure:
"""Create experiment comparison from checkbox selections"""
try:
if not selected_experiments:
fig = go.Figure()
fig.add_annotation(
text="Please select at least one experiment to compare",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="orange")
)
fig.update_layout(
title="No Experiments Selected",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
if not selected_metrics:
fig = go.Figure()
fig.add_annotation(
text="Please select at least one metric to compare",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="orange")
)
fig.update_layout(
title="No Metrics Selected",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
# Prepare dataframes for each selected experiment once
experiment_to_dataframe = {}
for experiment_id in selected_experiments:
try:
experiment_to_dataframe[experiment_id] = get_metrics_dataframe(experiment_id)
except Exception:
experiment_to_dataframe[experiment_id] = pd.DataFrame()
# Setup subplots: one subplot per selected metric
from plotly.subplots import make_subplots
num_metrics = len(selected_metrics)
num_columns = min(3, num_metrics)
num_rows = (num_metrics + num_columns - 1) // num_columns
fig = make_subplots(
rows=num_rows,
cols=num_columns,
subplot_titles=selected_metrics,
vertical_spacing=0.05,
horizontal_spacing=0.1
)
# Color palette for experiments (consistent colors across subplots)
try:
palette = px.colors.qualitative.Plotly
except Exception:
palette = [
'blue', 'red', 'green', 'orange', 'purple', 'brown',
'pink', 'gray', 'cyan', 'magenta'
]
experiment_to_color = {
exp_id: palette[idx % len(palette)] for idx, exp_id in enumerate(selected_experiments)
}
# Helper predicates (match logic used elsewhere in this file)
def _is_eval_metric(name: str) -> bool:
return name.startswith('eval_') or name.startswith('eval/')
def _is_system_metric(name: str) -> bool:
import re
if name in ("cpu_percent", "memory_percent"):
return True
return re.match(r"^gpu_\d+_(memory_allocated|memory_reserved|utilization)$", name) is not None
any_trace_added = False
for metric_index, metric_name in enumerate(selected_metrics):
row = (metric_index // num_columns) + 1
col = (metric_index % num_columns) + 1
subplot_has_data = False
for experiment_id, df in experiment_to_dataframe.items():
if df is None or df.empty or metric_name not in df.columns:
continue
# Build x/y based on metric type
try:
df_local = df.copy()
use_time_axis = _is_system_metric(metric_name)
if use_time_axis:
# Time axis: use timestamp → datetime
df_local['time'] = pd.to_datetime(df_local.get('timestamp', ''), errors='coerce')
if df_local['time'].isna().all():
df_local['time'] = range(1, len(df_local) + 1)
df_local.sort_values('time', inplace=True)
valid_mask = df_local[metric_name].notna()
x_values = df_local.loc[valid_mask, 'time'].tolist()
y_values = df_local.loc[valid_mask, metric_name].tolist()
else:
# Step axis: ensure a reasonable step column exists
if 'step' not in df_local or df_local['step'].nunique() <= 1:
for alternative in ['train/global_step', 'global_step', 'train/step']:
if alternative in df_local.columns and df_local[alternative].notna().any():
df_local['step'] = pd.to_numeric(df_local[alternative], errors='coerce')
break
if 'step' not in df_local.columns or df_local['step'].isna().all() or df_local['step'].nunique() <= 1:
df_local['step'] = range(1, len(df_local) + 1)
else:
df_local['step'] = pd.to_numeric(df_local.get('step', -1), errors='coerce').fillna(-1)
df_local.sort_values('step', inplace=True)
valid_mask = df_local[metric_name].notna()
x_values = df_local.loc[valid_mask, 'step'].tolist()
y_values = df_local.loc[valid_mask, metric_name].tolist()
except Exception:
# Fallback to naive arrays
valid_mask = df[metric_name].notna()
x_values = df.loc[valid_mask, 'step'].tolist() if 'step' in df.columns else list(range(1, len(df) + 1))
y_values = df.loc[valid_mask, metric_name].tolist() if metric_name in df.columns else []
if not x_values or not y_values:
continue
subplot_has_data = True
any_trace_added = True
color = experiment_to_color.get(experiment_id, 'blue')
fig.add_trace(
go.Scatter(
x=x_values,
y=y_values,
mode='lines+markers',
name=experiment_id,
line=dict(width=2, color=color),
marker=dict(size=4, color=color),
showlegend=True,
connectgaps=True if _is_eval_metric(metric_name) else False
),
row=row,
col=col
)
# Axis titles per subplot
try:
fig.update_xaxes(
title_text=("Time" if _is_system_metric(metric_name) else "Training Step"),
row=row,
col=col
)
fig.update_yaxes(title_text=metric_name, row=row, col=col)
except Exception:
pass
# If no experiment had data for this metric, annotate the subplot
if not subplot_has_data:
try:
fig.add_annotation(
text=f"No data for metric: {metric_name}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=12, color="gray"),
row=row, col=col
)
except Exception:
fig.add_annotation(
text=f"No data for metric: {metric_name}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=12, color="gray")
)
fig.update_layout(
title="Experiment Comparison",
height=max(350, 320 * num_rows),
plot_bgcolor='white',
paper_bgcolor='white',
hovermode='x unified',
legend=dict(orientation='h', yanchor='bottom', y=1.02, xanchor='right', x=1)
)
# Grid lines for all subplots
for r in range(1, num_rows + 1):
for c in range(1, num_columns + 1):
fig.update_xaxes(showgrid=True, gridwidth=1, gridcolor='lightgray', row=r, col=c)
fig.update_yaxes(showgrid=True, gridwidth=1, gridcolor='lightgray', row=r, col=c)
if not any_trace_added:
# Overall annotation if literally nothing to plot
fig = go.Figure()
fig.add_annotation(
text="No comparable data available for the selected experiments/metrics",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16, color="orange")
)
fig.update_layout(
title="No Data",
plot_bgcolor='white', paper_bgcolor='white'
)
return fig
except Exception as e:
logger.error(f"Error creating comparison from selection: {str(e)}")
fig = go.Figure()
fig.add_annotation(
text=f"Error creating comparison: {str(e)}",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=14, color="red")
)
return fig
def refresh_comparison_options() -> tuple:
"""Refresh the experiment and metric options for comparison"""
try:
# Get updated experiment choices
experiment_choices = get_experiment_dropdown_choices()
if experiment_choices == ["No experiments available"]:
experiment_choices = []
# Get available metrics from all experiments
all_experiments = list(trackio_space.experiments.keys())
available_metrics = get_available_metrics_for_experiments(all_experiments)
# Default to common metrics if available
default_metrics = []
common_metrics = ["loss", "accuracy", "learning_rate", "gpu_memory"]
for metric in common_metrics:
if metric in available_metrics:
default_metrics.append(metric)
# If no common metrics, use first few available
if not default_metrics and available_metrics:
default_metrics = available_metrics[:2]
return gr.CheckboxGroup(choices=experiment_choices, value=[]), gr.CheckboxGroup(choices=available_metrics, value=default_metrics)
except Exception as e:
logger.error(f"Error refreshing comparison options: {str(e)}")
return gr.CheckboxGroup(choices=[], value=[]), gr.CheckboxGroup(choices=["loss", "accuracy"], value=[])
# Create Gradio interface
with gr.Blocks(title="Trackio - Experiment Tracking", theme=gr.themes.Soft()) as demo:
gr.Markdown("# 🚀 Trackio Experiment Tracking & Monitoring")
gr.Markdown("Monitor and track your ML experiments with real-time visualization!")
with gr.Tabs():
# Dashboard Tab (NEW)
with gr.Tab("📊 Dashboard"):
gr.Markdown("### Comprehensive Experiment Dashboard")
gr.Markdown("Select an experiment to view all its data, plots, and information in one place.")
# Row 1: Experiment Selection
with gr.Row():
with gr.Column(scale=3):
# Experiment selection dropdown
experiment_dropdown = gr.Dropdown(
label="Select Experiment",
choices=get_experiment_dropdown_choices(),
value=get_experiment_dropdown_choices()[0] if get_experiment_dropdown_choices() and get_experiment_dropdown_choices()[0] != "No experiments available" else None,
info="Choose an experiment to view its dashboard"
)
with gr.Column(scale=1):
with gr.Row():
refresh_dropdown_btn = gr.Button("🔄 Refresh List", variant="secondary", size="sm")
refresh_dashboard_btn = gr.Button("🔄 Refresh Dashboard", variant="primary", size="sm")
# Row 2: All Metrics Plots
with gr.Row():
with gr.Column(scale=3):
with gr.Row():
gr.Markdown("### 📈 All Metrics Plots")
with gr.Row():
with gr.Column(scale=3):
dashboard_plots = gr.Plot(
label="Training Metrics",
container=True,
show_label=True,
elem_classes=["plot-container"]
)
# Row 3: Training Metrics Visualization Accordion
with gr.Row():
with gr.Accordion("📈 Training Metrics Visualization", open=False):
with gr.Row():
with gr.Column():
metric_dropdown = gr.Dropdown(
label="Metric to Plot",
choices=[
"loss", "accuracy", "learning_rate", "gpu_memory", "training_time",
"total_tokens", "truncated_tokens", "padding_tokens", "throughput", "step_time",
"batch_size", "seq_len", "token_acc", "train/gate_ortho", "train/center"
],
value="loss"
)
plot_btn = gr.Button("Create Plot", variant="primary")
test_plot_btn = gr.Button("Test Plot Rendering", variant="secondary")
with gr.Row():
dashboard_metric_plot = gr.Plot(
label="Training Metrics",
container=True,
show_label=True,
elem_classes=["plot-container"]
)
plot_btn.click(
create_metrics_plot,
inputs=[experiment_dropdown, metric_dropdown],
outputs=dashboard_metric_plot
)
test_plot_btn.click(
create_test_plot,
inputs=[],
outputs=dashboard_metric_plot
)
# Row 4: Accordion with Detailed Information
with gr.Row():
with gr.Accordion("📋 Experiment Details", open=False):
with gr.Tabs():
with gr.Tab("📋 Status"):
dashboard_status = gr.Textbox(
label="Experiment Status",
lines=8,
interactive=False
)
with gr.Tab("🔧 Parameters"):
dashboard_parameters = gr.Textbox(
label="Experiment Parameters",
lines=12,
interactive=False
)
with gr.Tab("📊 Metrics Summary"):
dashboard_metrics = gr.Textbox(
label="Metrics Summary",
lines=12,
interactive=False
)
with gr.Tab("📋 Complete Summary"):
dashboard_summary = gr.Textbox(
label="Full Experiment Summary",
lines=20,
interactive=False
)
# Connect the dashboard update function
experiment_dropdown.change(
update_dashboard,
inputs=[experiment_dropdown],
outputs=[dashboard_status, dashboard_parameters, dashboard_metrics, dashboard_plots, dashboard_summary]
)
refresh_dashboard_btn.click(
update_dashboard,
inputs=[experiment_dropdown],
outputs=[dashboard_status, dashboard_parameters, dashboard_metrics, dashboard_plots, dashboard_summary]
)
# Connect the metric plot update function
metric_dropdown.change(
update_dashboard_metric_plot,
inputs=[experiment_dropdown, metric_dropdown],
outputs=[dashboard_metric_plot]
)
refresh_dropdown_btn.click(
refresh_experiment_dropdown,
inputs=[],
outputs=[experiment_dropdown]
)
# Experiment Comparison Tab
with gr.Tab("📊 Experiment Comparison"):
gr.Markdown("### Compare Multiple Experiments")
gr.Markdown("Select experiments and metrics to compare from the available options below.")
# Selection controls
with gr.Row():
with gr.Column(scale=2):
gr.Markdown("### Available Experiments")
experiment_checkboxes = gr.CheckboxGroup(
label="Select Experiments to Compare",
choices=get_experiment_dropdown_choices(),
value=[],
info="Choose experiments to include in the comparison"
)
gr.Markdown("### Available Metrics")
metric_checkboxes = gr.CheckboxGroup(
label="Select Metrics to Compare",
choices=get_available_metrics_for_experiments(list(trackio_space.experiments.keys())),
value=["loss", "accuracy"],
info="Choose metrics to include in the comparison"
)
with gr.Row():
comparison_btn = gr.Button("Compare Selected", variant="primary")
refresh_options_btn = gr.Button("🔄 Refresh Options", variant="secondary")
with gr.Column(scale=1):
gr.Markdown("### Comparison Results")
gr.Markdown("The comparison will show subplots for the selected metrics across the selected experiments.")
# Comparison plots as subplots
comparison_plot = gr.Plot(
label="Experiment Comparison Dashboard",
container=True,
show_label=True,
elem_classes=["plot-container"]
)
comparison_btn.click(
create_experiment_comparison_from_selection,
inputs=[experiment_checkboxes, metric_checkboxes],
outputs=comparison_plot
)
refresh_options_btn.click(
refresh_comparison_options,
inputs=[],
outputs=[experiment_checkboxes, metric_checkboxes]
)
# Configuration Tab
with gr.Tab("⚙️ Configuration"):
gr.Markdown("### Configure HF Datasets Connection")
gr.Markdown("Set your Hugging Face token and dataset repository for persistent experiment storage.")
with gr.Row():
with gr.Column():
hf_token_input = gr.Textbox(
label="Hugging Face Token",
placeholder="hf_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
type="password",
info="Your HF token for dataset access (optional - will use environment variable if not set)"
)
dataset_repo_input = gr.Textbox(
label="Dataset Repository",
placeholder="your-username/your-dataset-name",
value="Tonic/trackio-experiments",
info="HF Dataset repository for experiment storage"
)
with gr.Row():
update_config_btn = gr.Button("Update Configuration", variant="primary")
test_connection_btn = gr.Button("Test Connection", variant="secondary")
create_repo_btn = gr.Button("Create Dataset", variant="success")
gr.Markdown("### Current Configuration")
current_config_output = gr.Textbox(
label="Status",
lines=10,
interactive=False,
value=f"📊 Dataset: {trackio_space.dataset_repo}\n🔑 HF Token: {'Set' if trackio_space.hf_token else 'Not set'}\n🛡️ Data Preservation: {'✅ Enabled' if trackio_space.dataset_manager else '⚠️ Legacy Mode'}\n📈 Experiments: {len(trackio_space.experiments)}\n📋 Available Experiments: {', '.join(list(trackio_space.experiments.keys())[:3])}{'...' if len(trackio_space.experiments) > 3 else ''}"
)
with gr.Column():
gr.Markdown("### Configuration Help")
gr.Markdown("""
**Getting Your HF Token:**
1. Go to [Hugging Face Settings](https://huggingface.co/settings/tokens)
2. Click "New token"
3. Give it a name (e.g., "Trackio Access")
4. Select "Write" permissions
5. Copy the token and paste it above
**Dataset Repository:**
- Format: `username/dataset-name`
- Examples: `tonic/trackio-experiments`, `your-username/my-experiments`
- Use "Create Dataset" button to create a new repository
**Environment Variables:**
You can also set these as environment variables:
- `HF_TOKEN`: Your Hugging Face token
- `TRACKIO_DATASET_REPO`: Dataset repository
**Data Preservation:**
- ✅ **Enabled**: All experiment data is preserved when adding/updating experiments
- ⚠️ **Legacy Mode**: Data preservation not guaranteed (fallback mode)
- Data preservation requires the dataset management utilities to be available
**Actions:**
- **Update Configuration**: Apply new settings and reload experiments
- **Test Connection**: Verify access to the dataset repository
- **Create Dataset**: Create a new dataset repository if it doesn't exist
""")
# Experiment Management Accordion
with gr.Accordion("🔧 Experiment Management", open=False):
with gr.Tabs():
# Create Experiment Tab
with gr.Tab("Create Experiment"):
gr.Markdown("### Create a New Experiment")
with gr.Row():
with gr.Column():
create_exp_name = gr.Textbox(
label="Experiment Name",
placeholder="my_smollm3_finetune",
value="smollm3_finetune"
)
create_exp_description = gr.Textbox(
label="Description",
placeholder="Fine-tuning SmolLM3 model on custom dataset",
value="SmolLM3 fine-tuning experiment"
)
create_exp_btn = gr.Button("Create Experiment", variant="primary")
with gr.Column():
create_exp_output = gr.Textbox(
label="Result",
lines=5,
interactive=False
)
create_exp_btn.click(
create_experiment_interface,
inputs=[create_exp_name, create_exp_description],
outputs=[create_exp_output, experiment_dropdown]
)
# Log Metrics Tab
with gr.Tab("Log Metrics"):
gr.Markdown("### Log Training Metrics")
with gr.Row():
with gr.Column():
log_metrics_exp_id = gr.Textbox(
label="Experiment ID",
placeholder="exp_20231201_143022"
)
log_metrics_json = gr.Textbox(
label="Metrics (JSON)",
placeholder='{"loss": 0.5, "accuracy": 0.85, "learning_rate": 2e-5}',
value='{"loss": 0.5, "accuracy": 0.85, "learning_rate": 2e-5, "gpu_memory": 22.5}'
)
log_metrics_step = gr.Textbox(
label="Step (optional)",
placeholder="100"
)
log_metrics_btn = gr.Button("Log Metrics", variant="primary")
with gr.Column():
log_metrics_output = gr.Textbox(
label="Result",
lines=5,
interactive=False
)
log_metrics_btn.click(
log_metrics_interface,
inputs=[log_metrics_exp_id, log_metrics_json, log_metrics_step],
outputs=log_metrics_output
)
# Log Parameters Tab
with gr.Tab("Log Parameters"):
gr.Markdown("### Log Experiment Parameters")
with gr.Row():
with gr.Column():
log_params_exp_id = gr.Textbox(
label="Experiment ID",
placeholder="exp_20231201_143022"
)
log_params_json = gr.Textbox(
label="Parameters (JSON)",
placeholder='{"learning_rate": 2e-5, "batch_size": 4}',
value='{"learning_rate": 3.5e-6, "batch_size": 8, "model_name": "HuggingFaceTB/SmolLM3-3B", "max_iters": 18000, "mixed_precision": "bf16"}'
)
log_params_btn = gr.Button("Log Parameters", variant="primary")
with gr.Column():
log_params_output = gr.Textbox(
label="Result",
lines=5,
interactive=False
)
log_params_btn.click(
log_parameters_interface,
inputs=[log_params_exp_id, log_params_json],
outputs=log_params_output
)
# View Experiments Tab
with gr.Tab("View Experiments"):
gr.Markdown("### View Experiment Details")
with gr.Row():
with gr.Column():
view_exp_id = gr.Textbox(
label="Experiment ID",
placeholder="exp_20231201_143022"
)
view_btn = gr.Button("View Experiment", variant="primary")
list_btn = gr.Button("List All Experiments", variant="secondary")
with gr.Column():
view_output = gr.Textbox(
label="Experiment Details",
lines=20,
interactive=False
)
view_btn.click(
get_experiment_details,
inputs=[view_exp_id],
outputs=view_output
)
list_btn.click(
list_experiments_interface,
inputs=[],
outputs=view_output
)
# Update Status Tab
with gr.Tab("Update Status"):
gr.Markdown("### Update Experiment Status")
with gr.Row():
with gr.Column():
status_exp_id = gr.Textbox(
label="Experiment ID",
placeholder="exp_20231201_143022"
)
status_dropdown = gr.Dropdown(
label="Status",
choices=["running", "completed", "failed", "paused"],
value="running"
)
update_status_btn = gr.Button("Update Status", variant="primary")
with gr.Column():
status_output = gr.Textbox(
label="Result",
lines=3,
interactive=False
)
update_status_btn.click(
update_experiment_status_interface,
inputs=[status_exp_id, status_dropdown],
outputs=status_output
)
# Demo Data Tab
with gr.Tab("Demo Data"):
gr.Markdown("### Generate Demo Training Data")
gr.Markdown("Use this to simulate training data for testing the interface")
with gr.Row():
with gr.Column():
demo_exp_id = gr.Textbox(
label="Experiment ID",
placeholder="exp_20231201_143022"
)
demo_btn = gr.Button("Generate Demo Data", variant="primary")
create_demo_btn = gr.Button("Create Demo Experiment", variant="secondary")
with gr.Column():
demo_output = gr.Textbox(
label="Result",
lines=5,
interactive=False
)
demo_btn.click(
simulate_training_data,
inputs=[demo_exp_id],
outputs=[demo_output, dashboard_status, dashboard_parameters, dashboard_metrics, dashboard_plots, dashboard_summary]
)
create_demo_btn.click(
create_demo_experiment,
inputs=[],
outputs=[demo_output, experiment_dropdown]
)
update_config_btn.click(
update_trackio_config,
inputs=[hf_token_input, dataset_repo_input],
outputs=current_config_output
)
test_connection_btn.click(
test_dataset_connection,
inputs=[hf_token_input, dataset_repo_input],
outputs=current_config_output
)
create_repo_btn.click(
create_dataset_repository,
inputs=[hf_token_input, dataset_repo_input],
outputs=current_config_output
)
# Launch the app
if __name__ == "__main__":
demo.launch(show_error=True) |