Knowledge

File:Estimated training cost of some AI models - 2024 AI index.jpg

Source 📝

38: 265: 287: 201: 293: 298: 94: 663: 148:
It shows the rapid increase in the computing used to train large language models. It also shows that the training cost of the best closed large language models seems much higher than the training cost of the best open-weight ones, leading to higher performance. The training cost of models like GPT-4
264: 680:
Uploaded a work by Stanford Institute for Human-Centered Artificial Intelligence (permission obtained by email from the AI index research manager) from https://aiindex.stanford.edu/report/#individual-chapters with
339:– You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. 60: 56: 52: 48: 42: 105: 73: 205: 223:
The Wikimedia Foundation has received an e-mail confirming that the copyright holder has approved publication under the terms mentioned on this page. This correspondence has been
239: 608: 718: 149:
is not publicly known, so this is just an estimate. The data is from Epoch in 2023, and the chart is from Stanford University's 2024 AI index.
248: 217: 188: 538: 502: 662: 657: 179:
Stanford Institute for Human-Centered Artificial Intelligence (permission obtained by email from the AI index research manager)
64: 37: 419: 346: 592: 486: 457: 439: 255: 558: 522: 432: 403: 234: 230: 221:, you do not need to request permission as long as you follow any licensing requirements mentioned on this page. 308: 464: 345:– If you remix, transform, or build upon the material, you must distribute your contributions under the 700: 305: 256:
https://ticket.wikimedia.org/otrs/index.pl?Action=AgentTicketZoom&TicketNumber=2024060510011799
113: 695: 212: 168: 673: 100: 158: 247:
If you have questions about the archived correspondence, please use the
738: 630:
Click on a date/time to view the file as it appeared at that time.
728: 195: 88: 138:
Estimated training cost of some AI models - 2024 AI index.jpg
237:. The correspondence is available to trusted volunteers as 215:
and may be used by anyone for any purpose. If you wish to
539:
Creative Commons Attribution-ShareAlike 4.0 International
465:
https://aiindex.stanford.edu/report/#individual-chapters
169:
https://aiindex.stanford.edu/report/#individual-chapters
387:
Add a one-line explanation of what this file represents
78:(2,560 × 1,361 pixels, file size: 147 KB, MIME type: 132: 112:
Commons is a freely licensed media file repository.
356:https://creativecommons.org/licenses/by-sa/4.0 93: 362:Creative Commons Attribution-Share Alike 4.0 286: 8: 322:– to copy, distribute and transmit the work 632: 383: 711:The following other wikis use this file: 309:Attribution-Share Alike 4.0 International 693: 606: 590: 572: 556: 536: 520: 500: 484: 450: 417: 401: 398: 379: 372: 261:Find other files from the same ticket: 691:The following 2 pages use this file: 200: 7: 231:Volunteer Response Team (VRT) member 621: 297: 292: 225: 392: 386: 283: 142: 125: 71: 391: 377: 333:Under the following conditions: 304:This file is licensed under the 296: 291: 285: 263: 199: 92: 31: 21: 374: 143: 14: 373: 26: 1: 593:Wikimedia VRTS ticket number 399:Items portrayed in this file 735:Usage on kaa.wikipedia.org 715:Usage on ckb.wikipedia.org 757: 725:Usage on hu.wikipedia.org 347:same or compatible license 622: 16: 433:language of work or name 376: 313: 240:ticket #2024060510011799 198: 99:This is a file from the 679: 183: 175: 164: 157: 154: 135: 103:. Information from its 106:description page there 41:Size of this preview: 701:Large language model 171:(chapter 1, image 3) 65:2,560 × 1,361 pixels 328:– to adapt the work 47:Other resolutions: 729:Nagy nyelvi modell 719:مۆدێلی زمانی گەورە 658:20:52, 6 June 2024 235:permission archive 233:and stored in our 61:1,280 × 681 pixels 57:1,024 × 544 pixels 707:Global file usage 684: 523:copyright license 385: 276: 275: 272: 271: 262: 258: 252: 192: 189:Reusing this file 121: 120: 101:Wikimedia Commons 32:Global file usage 748: 739:Úlken til modeli 696:Foundation model 670: 609:2024060510011799 487:copyright status 467: 462: 442: 437: 369: 366: 363: 360: 357: 349:as the original. 306:Creative Commons 300: 299: 295: 294: 289: 288: 267: 260: 253: 246: 227: 218:use this content 207: 203: 202: 196: 186: 160: 147: 139: 133: 117: 96: 95: 89: 83: 81: 68: 53:640 × 340 pixels 49:320 × 170 pixels 43:800 × 425 pixels 756: 755: 751: 750: 749: 747: 746: 745: 705: 685: 676: 668: 624: 623: 620: 619: 618: 617: 616: 615: 614: 613: 611: 599: 598: 597: 595: 584: 583: 582: 581: 580: 579: 578: 577: 575: 565: 564: 563: 561: 550: 549: 548: 547: 546: 545: 544: 543: 541: 529: 528: 527: 525: 514: 513: 512: 511: 510: 509: 508: 507: 505: 493: 492: 491: 489: 478: 477: 476: 475: 474: 473: 472: 471: 470: 469: 468: 463: 460: 447: 446: 445: 444: 443: 438: 435: 426: 425: 424: 422: 410: 409: 408: 406: 390: 389: 388: 371: 370: 367: 364: 361: 358: 355: 354: 312: 301: 282: 277: 249:VRT noticeboard 150: 137: 130: 123: 122: 111: 110: 109:is shown below. 85: 79: 77: 70: 69: 46: 12: 11: 5: 754: 752: 744: 743: 742: 741: 733: 732: 731: 723: 722: 721: 709: 708: 704: 703: 698: 689: 688: 683: 682: 678: 674: 671: 667:2,560 × 1,361 665: 660: 655: 651: 650: 647: 644: 641: 638: 635: 628: 627: 612: 607: 605: 604: 603: 602: 601: 600: 596: 591: 589: 588: 587: 586: 585: 576: 573: 571: 570: 569: 568: 567: 566: 562: 557: 555: 554: 553: 552: 551: 542: 537: 535: 534: 533: 532: 531: 530: 526: 521: 519: 518: 517: 516: 515: 506: 501: 499: 498: 497: 496: 495: 494: 490: 485: 483: 482: 481: 480: 479: 456: 455: 454: 453: 452: 449: 448: 431: 430: 429: 428: 427: 423: 418: 416: 415: 414: 413: 412: 411: 407: 402: 400: 397: 396: 395: 394: 393: 382: 381: 378: 375: 353: 352: 351: 350: 340: 331: 330: 329: 323: 316:You are free: 303: 302: 284: 281: 278: 274: 273: 270: 269: 209: 193: 181: 180: 177: 173: 172: 166: 162: 161: 156: 152: 151: 140: 131: 129: 126: 124: 119: 118: 97: 87: 86: 40: 36: 35: 34: 29: 24: 19: 13: 10: 9: 6: 4: 3: 2: 753: 740: 737: 736: 734: 730: 727: 726: 724: 720: 717: 716: 714: 713: 712: 706: 702: 699: 697: 694: 692: 686: 677: 672: 666: 664: 661: 659: 656: 653: 652: 648: 645: 642: 639: 636: 634: 633: 631: 625: 610: 594: 574:15 April 2024 560: 540: 524: 504: 488: 466: 459: 441: 434: 421: 405: 359:CC BY-SA 4.0 348: 344: 341: 338: 335: 334: 332: 327: 324: 321: 318: 317: 315: 314: 310: 307: 290: 279: 268: 266: 259: 257: 254:Ticket link: 250: 244: 242: 241: 236: 232: 228: 220: 219: 214: 211:This work is 210: 208: 197: 194: 190: 185: 182: 178: 174: 170: 167: 163: 159:15 April 2024 153: 146: 141: 134: 127: 115: 108: 107: 102: 98: 91: 90: 84: 75: 74:Original file 66: 62: 58: 54: 50: 44: 39: 33: 30: 28: 25: 23: 20: 18: 15: 710: 690: 681:UploadWizard 629: 626:File history 342: 336: 325: 319: 245: 238: 224: 222: 216: 204: 144: 114:You can help 104: 72: 22:File history 503:copyrighted 343:share alike 337:attribution 136:Description 687:File usage 643:Dimensions 184:Permission 80:image/jpeg 27:File usage 640:Thumbnail 637:Date/Time 559:inception 451:Reference 280:Licensing 145:English: 675:Alenoach 669:(147 KB) 380:Captions 326:to remix 320:to share 311:license. 226:reviewed 654:current 649:Comment 440:English 420:diagram 404:depicts 384:English 128:Summary 76:‎ 206:  176:Author 165:Source 229:by a 646:User 368:true 365:true 213:free 155:Date 17:File 458:URL 461:: 436:: 243:. 63:| 59:| 55:| 51:| 45:. 251:. 191:) 187:( 116:. 82:) 67:.

Index

File
File history
File usage
Global file usage
File:Estimated training cost of some AI models - 2024 AI index.jpg
800 × 425 pixels
320 × 170 pixels
640 × 340 pixels
1,024 × 544 pixels
1,280 × 681 pixels
2,560 × 1,361 pixels
Original file
Wikimedia Commons
description page there
You can help
https://aiindex.stanford.edu/report/#individual-chapters
Reusing this file
free
use this content
Volunteer Response Team (VRT) member
permission archive
ticket #2024060510011799
VRT noticeboard
https://ticket.wikimedia.org/otrs/index.pl?Action=AgentTicketZoom&TicketNumber=2024060510011799
SDC query (SPARQL)
Creative Commons
Attribution-Share Alike 4.0 International
same or compatible license
depicts
diagram

Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.