kreas commited on
Commit
f627450
1 Parent(s): 31b742e

Upload /gemma-2b/nf4_batch_size_1_sq_len_256_new_tokens_256/benchmark_report.json with huggingface_hub

Browse files
gemma-2b/nf4_batch_size_1_sq_len_256_new_tokens_256/benchmark_report.json CHANGED
@@ -2,410 +2,410 @@
2
  "prefill": {
3
  "memory": {
4
  "unit": "MB",
5
- "max_ram": 1079.635968,
6
- "max_vram": 3193.962496,
7
  "max_reserved": 2709.520384,
8
  "max_allocated": 2546.903552
9
  },
10
  "latency": {
11
  "unit": "s",
12
- "mean": 0.025713854238122807,
13
- "stdev": 0.00030360055341363323,
14
  "values": [
15
- 0.028421695709228516,
16
- 0.026586111068725587,
17
- 0.02635897636413574,
18
- 0.026595327377319337,
19
- 0.02615616035461426,
20
- 0.02615091133117676,
21
- 0.026216447830200194,
22
- 0.02633830451965332,
23
- 0.02635775947570801,
24
- 0.028296192169189452,
25
- 0.028580863952636718,
26
- 0.02812723159790039,
27
- 0.026614784240722656,
28
- 0.025801727294921875,
29
- 0.025643007278442383,
30
- 0.02571161651611328,
31
- 0.02567987251281738,
32
- 0.02573004722595215,
33
- 0.025627647399902344,
34
- 0.02555392074584961,
35
- 0.02555904006958008,
36
- 0.02552524757385254,
37
- 0.025627647399902344,
38
- 0.025594879150390625,
39
- 0.025639936447143553,
40
- 0.025636863708496094,
41
- 0.025624576568603515,
42
- 0.025669631958007814,
43
- 0.025668607711791993,
44
- 0.026396671295166017,
45
- 0.025655296325683592,
46
- 0.025596927642822266,
47
- 0.025604095458984375,
48
- 0.025593856811523437,
49
- 0.02571571159362793,
50
- 0.025676799774169923,
51
- 0.025693183898925782,
52
- 0.025657567977905273,
53
- 0.025619455337524414,
54
- 0.025572351455688477,
55
- 0.025641984939575195,
56
- 0.025677824020385744,
57
- 0.025586687088012695,
58
- 0.025771007537841797,
59
- 0.02571059226989746,
60
- 0.02568191909790039,
61
- 0.025643007278442383,
62
- 0.0256790714263916,
63
- 0.025750528335571288,
64
- 0.025626623153686523,
65
- 0.02572390365600586,
66
- 0.025693183898925782,
67
- 0.02567884826660156,
68
- 0.025667680740356445,
69
- 0.025677824020385744,
70
- 0.025693183898925782,
71
- 0.025705472946166992,
72
- 0.02569830322265625,
73
- 0.025693183898925782,
74
- 0.02565017509460449,
75
- 0.025603071212768554,
76
- 0.025639936447143553,
77
- 0.025669631958007814,
78
- 0.026003456115722655,
79
- 0.025650432586669922,
80
- 0.025627647399902344,
81
- 0.025637887954711915,
82
- 0.025620479583740235,
83
- 0.025596927642822266,
84
- 0.02570854377746582,
85
- 0.025680896759033203,
86
  0.025646080017089845,
87
- 0.02559292793273926,
88
- 0.025628671646118165,
89
- 0.025656320571899413,
90
- 0.025630720138549806,
91
- 0.025626623153686523,
92
- 0.025656320571899413,
93
- 0.025659391403198242,
94
- 0.025604095458984375,
95
- 0.025623552322387694,
96
- 0.025882623672485353,
97
- 0.025602048873901367,
98
- 0.025581567764282227,
99
- 0.02564019203186035,
100
- 0.02571468734741211,
101
- 0.025657344818115234,
102
- 0.025643007278442383,
103
- 0.025655296325683592,
104
- 0.025636863708496094,
105
- 0.025693183898925782,
106
- 0.025605119705200196,
107
- 0.025582592010498048,
108
- 0.025683967590332032,
109
- 0.025585887908935546,
110
- 0.025625600814819335,
111
- 0.025606143951416017,
112
- 0.025663488388061522,
113
- 0.025602048873901367,
114
- 0.025652223587036133,
115
- 0.02555084800720215,
116
- 0.025608415603637694,
117
- 0.025586687088012695,
118
- 0.025596927642822266,
119
  0.025648128509521483,
120
- 0.025610240936279297,
121
- 0.025595903396606445,
122
  0.025654272079467775,
123
  0.025622528076171876,
124
- 0.025627647399902344,
125
- 0.02561561584472656,
126
- 0.025592832565307616,
127
- 0.025609216690063476,
128
- 0.025610240936279297,
129
- 0.025593856811523437,
130
- 0.025595903396606445,
131
- 0.025646080017089845,
132
- 0.025613311767578126,
133
- 0.025671680450439452,
134
- 0.025647104263305662,
135
- 0.025613311767578126,
136
- 0.025586687088012695,
137
- 0.025623552322387694,
138
- 0.025612287521362305,
139
- 0.025671680450439452,
140
- 0.025602048873901367,
141
- 0.02572287940979004,
142
- 0.02569113540649414,
143
- 0.025694303512573242,
144
- 0.0257126407623291,
145
- 0.02565836715698242,
146
- 0.02555392074584961,
147
- 0.025644031524658203,
148
  0.025659391403198242,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
149
  0.02571468734741211,
150
- 0.025626623153686523,
151
- 0.02571161651611328,
152
- 0.02569545555114746,
153
- 0.02567884826660156,
154
- 0.025846784591674804,
155
- 0.025596927642822266,
156
- 0.026058752059936522,
157
- 0.025607168197631838,
158
- 0.025639936447143553,
159
- 0.02568294334411621,
160
- 0.025613311767578126,
161
- 0.025667648315429687,
162
  0.0257126407623291,
163
- 0.02565017509460449,
164
- 0.02567475128173828,
165
  0.025683967590332032,
166
- 0.02555904006958008,
 
 
 
 
 
 
167
  0.025686016082763673,
168
- 0.025689088821411132,
169
- 0.025684991836547853,
170
- 0.025671808242797852,
171
- 0.025662464141845705,
172
- 0.025638912200927736,
173
- 0.025580543518066406,
174
- 0.025652223587036133,
175
- 0.02612735939025879,
176
- 0.02568294334411621,
177
- 0.025653247833251954,
178
- 0.02570751953125,
179
- 0.025677824020385744,
180
- 0.025659391403198242,
181
- 0.025663488388061522,
182
- 0.025641984939575195,
183
- 0.025634815216064453,
184
- 0.025717760086059572,
185
- 0.025630720138549806,
186
- 0.02569113540649414,
187
- 0.02569113540649414,
188
- 0.02567987251281738,
189
- 0.025661535263061523,
190
- 0.02561440086364746,
191
- 0.025627647399902344,
192
- 0.025586687088012695,
193
- 0.025634815216064453,
194
- 0.025628671646118165,
195
- 0.02571059226989746,
196
- 0.025664512634277343,
197
  0.02569523239135742,
 
 
198
  0.02574131202697754,
199
- 0.025694368362426757,
200
- 0.025647104263305662,
201
- 0.02572185516357422,
202
- 0.025632768630981444,
203
- 0.025616384506225585,
204
- 0.025673728942871094,
205
- 0.02569011116027832,
206
- 0.02570240020751953,
207
- 0.025742399215698243,
208
- 0.025635839462280274,
209
- 0.025687135696411133,
210
- 0.025667583465576172,
211
- 0.025686016082763673,
212
- 0.025680896759033203,
213
  0.02574131202697754,
214
- 0.02572800064086914,
215
- 0.02569113540649414,
216
- 0.02568720054626465,
217
- 0.025636863708496094,
218
- 0.02567065620422363,
 
 
 
 
 
219
  0.025667583465576172,
220
- 0.025630720138549806,
221
- 0.0257259521484375,
222
- 0.025627647399902344,
223
- 0.025633792877197265,
224
- 0.025618431091308593,
225
- 0.025654272079467775,
226
- 0.02568806457519531,
227
- 0.02569113540649414,
228
- 0.025598976135253908,
229
- 0.025604095458984375,
230
- 0.025656320571899413,
231
- 0.025654272079467775,
232
- 0.02568294334411621,
233
- 0.025671680450439452,
234
  0.025667583465576172,
235
- 0.026065919876098635,
236
- 0.02572083282470703,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
237
  0.025649152755737304,
238
- 0.02569830322265625,
239
- 0.02560620880126953,
240
- 0.025672704696655273,
241
- 0.02571366310119629,
242
- 0.025568256378173827,
243
- 0.02568191909790039,
244
  0.025643007278442383,
245
- 0.025689088821411132,
246
- 0.025793535232543945,
247
- 0.02571161651611328,
248
- 0.025796607971191408,
249
- 0.025701375961303712,
250
- 0.025634815216064453,
251
- 0.02570240020751953,
252
- 0.025609439849853515,
253
- 0.02589388847351074,
254
- 0.025785343170166015,
255
- 0.025668607711791993,
256
- 0.025644031524658203,
 
257
  0.025622528076171876,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
258
  0.025585664749145507,
259
- 0.025675775527954102,
260
- 0.025602048873901367,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
261
  0.025624576568603515,
 
 
 
 
 
262
  0.025601024627685546,
263
- 0.025605119705200196,
264
- 0.02571673583984375,
 
 
 
 
 
 
265
  0.025631744384765624,
266
- 0.02566655921936035,
267
- 0.025617408752441406,
268
- 0.025604095458984375,
269
- 0.025689088821411132,
270
- 0.025632768630981444,
271
- 0.025591808319091795,
 
 
 
 
 
 
 
 
 
 
 
272
  0.025571327209472656,
273
- 0.02564918327331543,
274
- 0.025643007278442383,
275
- 0.025603071212768554,
276
- 0.025693183898925782,
277
- 0.025659391403198242,
278
- 0.025603071212768554,
279
- 0.02569011116027832,
280
- 0.02572185516357422,
281
- 0.02575564765930176,
282
- 0.025653247833251954,
283
  0.025648128509521483,
284
- 0.02569036865234375,
285
  0.025645055770874024,
286
- 0.025637887954711915,
287
- 0.025646080017089845,
288
- 0.025627647399902344,
289
- 0.025652223587036133,
290
- 0.025593856811523437,
291
- 0.025618431091308593,
292
  0.025585664749145507,
293
- 0.025641984939575195,
294
- 0.025657344818115234,
295
- 0.02572185516357422,
296
- 0.025625823974609375,
297
- 0.025602048873901367,
 
 
 
 
 
 
298
  0.025587711334228515,
299
- 0.025664512634277343,
300
- 0.0257126407623291,
301
- 0.025638912200927736,
 
 
 
 
 
 
 
 
 
 
 
302
  0.025635839462280274,
303
- 0.025630720138549806,
304
- 0.02565836715698242,
305
- 0.025769983291625977,
306
- 0.02563279914855957,
307
- 0.025624576568603515,
308
- 0.025639936447143553,
309
- 0.025686016082763673,
310
- 0.025723968505859375,
311
- 0.025645055770874024,
312
- 0.02567065620422363,
313
- 0.02613555145263672,
314
- 0.025668607711791993,
315
- 0.025758720397949218,
316
- 0.025665536880493164,
317
- 0.0256911678314209,
318
- 0.02567884826660156,
319
- 0.025742431640625,
320
- 0.02572185516357422,
321
- 0.025647104263305662,
322
- 0.02570444869995117,
323
- 0.025646080017089845,
324
- 0.02574847984313965,
325
- 0.025677824020385744,
326
  0.025629695892333985,
327
- 0.025584640502929686,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
328
  0.025552896499633788,
329
- 0.02568704032897949,
330
- 0.02569011116027832,
331
- 0.025644031524658203,
332
- 0.0261396484375,
333
- 0.02558361625671387,
334
- 0.025628671646118165,
335
- 0.025761791229248047,
336
- 0.025637887954711915,
337
- 0.025663488388061522,
 
 
 
 
 
 
 
 
 
 
338
  0.025625600814819335,
339
- 0.025686016082763673,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
340
  0.02570854377746582,
 
 
 
 
 
 
 
 
 
341
  0.025614336013793947,
342
- 0.025625696182250978,
343
- 0.025639936447143553,
 
344
  0.0257259521484375,
345
- 0.025669631958007814,
346
- 0.025609216690063476,
347
- 0.025637887954711915,
348
- 0.025607168197631838,
349
- 0.02565017509460449,
350
- 0.02574028778076172,
351
- 0.025668607711791993,
352
- 0.02592051124572754,
353
- 0.025672704696655273,
354
- 0.025652223587036133,
355
- 0.025798656463623046,
356
- 0.025689088821411132,
357
- 0.02569011116027832,
358
- 0.025635839462280274,
359
- 0.02568294334411621,
360
- 0.02572185516357422,
361
- 0.0257127685546875,
362
- 0.025641984939575195,
363
- 0.025659391403198242,
364
- 0.025610240936279297,
365
- 0.02573619270324707,
366
- 0.025610240936279297,
367
- 0.02572697639465332,
368
- 0.025767936706542968,
369
- 0.025659391403198242,
370
- 0.025613311767578126,
371
- 0.025610240936279297,
372
- 0.025653247833251954,
373
- 0.025717760086059572,
374
- 0.02572083282470703,
375
- 0.025661439895629884,
376
- 0.025648128509521483,
377
- 0.025661439895629884,
378
  0.025660415649414063,
379
- 0.025777151107788086,
380
- 0.025659391403198242,
381
- 0.025660480499267578,
382
- 0.025632768630981444,
383
- 0.02572902488708496,
384
- 0.02573423957824707,
385
- 0.025578496932983398,
386
- 0.025623552322387694,
387
- 0.025639936447143553,
388
- 0.025632768630981444,
389
- 0.025667583465576172,
390
- 0.025662464141845705,
391
- 0.025652223587036133,
392
- 0.026096639633178712,
393
- 0.025682016372680663,
394
- 0.02569011116027832,
395
- 0.02570342445373535,
396
- 0.02572800064086914,
397
- 0.025639936447143553,
398
- 0.025659391403198242,
399
- 0.02556211280822754,
400
- 0.025640960693359374,
401
- 0.02570240020751953,
402
- 0.025785343170166015,
403
- 0.025668607711791993
404
  ]
405
  },
406
  "throughput": {
407
  "unit": "tokens/s",
408
- "value": 9955.722608882954
409
  },
410
  "energy": null,
411
  "efficiency": null
@@ -413,549 +413,549 @@
413
  "decode": {
414
  "memory": {
415
  "unit": "MB",
416
- "max_ram": 1079.635968,
417
  "max_vram": 3472.883712,
418
  "max_reserved": 2988.4416,
419
  "max_allocated": 2813.773312
420
  },
421
  "latency": {
422
  "unit": "s",
423
- "mean": 5.112615415573122,
424
- "stdev": 0.008767993927004447,
425
  "values": [
426
- 5.121383409500126,
427
- 5.103847421646117
428
  ]
429
  },
430
  "throughput": {
431
  "unit": "tokens/s",
432
- "value": 49.87662463780578
433
  },
434
  "energy": null,
435
  "efficiency": null
436
- },
437
- "per_token": {
438
- "memory": null,
439
- "latency": {
440
- "unit": "s",
441
- "mean": 0.02004947221793381,
442
- "stdev": 0.00010631814525496304,
443
- "values": [
444
- 0.02019327926635742,
445
- 0.02003660774230957,
446
- 0.02008166313171387,
447
- 0.020094175338745118,
448
- 0.019956512451171873,
449
- 0.020099327087402342,
450
- 0.020105983734130858,
451
- 0.020489215850830078,
452
- 0.020000768661499024,
453
- 0.020092927932739257,
454
- 0.019938304901123048,
455
- 0.020049983978271485,
456
- 0.02016864013671875,
457
- 0.020075519561767577,
458
- 0.02008576011657715,
459
- 0.0200130558013916,
460
- 0.019970304489135744,
461
- 0.020060928344726562,
462
- 0.020075519561767577,
463
- 0.020121599197387697,
464
- 0.020136959075927736,
465
- 0.020123647689819335,
466
- 0.02001919937133789,
467
- 0.02003558349609375,
468
- 0.019970048904418947,
469
- 0.02002943992614746,
470
- 0.02007142448425293,
471
- 0.020039680480957032,
472
- 0.020165632247924805,
473
- 0.019990528106689453,
474
- 0.020092927932739257,
475
- 0.020084735870361328,
476
- 0.020084735870361328,
477
- 0.02008166313171387,
478
- 0.020212736129760742,
479
- 0.020084735870361328,
480
- 0.020006912231445313,
481
- 0.019965951919555663,
482
- 0.02007347106933594,
483
- 0.019989696502685547,
484
- 0.02000467109680176,
485
- 0.02006425666809082,
486
- 0.02006425666809082,
487
- 0.02007859230041504,
488
- 0.02004787254333496,
489
- 0.019944448471069336,
490
- 0.020032512664794923,
491
- 0.019960832595825196,
492
- 0.02005504035949707,
493
- 0.02002739143371582,
494
- 0.019916799545288084,
495
- 0.01990553665161133,
496
- 0.01992192077636719,
497
- 0.019951616287231445,
498
- 0.019959808349609375,
499
- 0.019978239059448243,
500
- 0.019946495056152345,
501
- 0.019994623184204103,
502
- 0.020122623443603514,
503
- 0.020147199630737304,
504
- 0.02004787254333496,
505
- 0.02000588798522949,
506
- 0.02003046417236328,
507
- 0.020024320602416993,
508
- 0.020136959075927736,
509
- 0.02003865623474121,
510
- 0.019998720169067383,
511
- 0.020025375366210938,
512
- 0.019988447189331054,
513
- 0.020172800064086914,
514
- 0.0200130558013916,
515
- 0.020170751571655272,
516
- 0.02003558349609375,
517
- 0.020132896423339843,
518
- 0.020011999130249022,
519
- 0.019976192474365235,
520
- 0.019987455368041994,
521
- 0.020000768661499024,
522
- 0.019991552352905274,
523
- 0.019986656188964842,
524
- 0.019947296142578126,
525
- 0.020145151138305666,
526
- 0.020116479873657226,
527
- 0.020114431381225584,
528
- 0.020496383666992187,
529
- 0.020050943374633787,
530
- 0.020039680480957032,
531
- 0.020177024841308594,
532
- 0.020014976501464842,
533
- 0.020021440505981446,
534
- 0.02000467109680176,
535
- 0.020114431381225584,
536
- 0.020123647689819335,
537
- 0.02003763198852539,
538
- 0.020025344848632814,
539
- 0.019994623184204103,
540
- 0.02003763198852539,
541
- 0.02001817512512207,
542
- 0.0200263671875,
543
- 0.020057088851928712,
544
- 0.020160512924194338,
545
- 0.020106239318847655,
546
- 0.02020966339111328,
547
- 0.020006975173950194,
548
- 0.020070335388183595,
549
- 0.020059232711791993,
550
- 0.020120479583740233,
551
- 0.020117504119873047,
552
- 0.020555776596069338,
553
- 0.020108287811279296,
554
- 0.02012057685852051,
555
- 0.020002912521362305,
556
- 0.019954591751098632,
557
- 0.02001919937133789,
558
- 0.02025369644165039,
559
- 0.020139007568359374,
560
- 0.020140031814575195,
561
- 0.02008166313171387,
562
- 0.019993600845336915,
563
- 0.02006447982788086,
564
- 0.020110111236572265,
565
- 0.02024345588684082,
566
- 0.02012774467468262,
567
- 0.02011136054992676,
568
- 0.019994623184204103,
569
- 0.01997123146057129,
570
- 0.019997535705566408,
571
- 0.02000486373901367,
572
- 0.020031488418579102,
573
- 0.020060319900512696,
574
- 0.020019039154052735,
575
- 0.02000611114501953,
576
- 0.02008041572570801,
577
- 0.020135936737060548,
578
- 0.020229120254516602,
579
- 0.02010316848754883,
580
- 0.020126720428466797,
581
- 0.0200447998046875,
582
- 0.020063232421875,
583
- 0.02007347106933594,
584
- 0.020065280914306642,
585
- 0.020031488418579102,
586
- 0.020118528366088868,
587
- 0.02023526382446289,
588
- 0.02018611145019531,
589
- 0.020069408416748046,
590
- 0.020046815872192383,
591
- 0.020039871215820314,
592
- 0.020032320022583008,
593
- 0.020123647689819335,
594
- 0.02001408004760742,
595
- 0.019998720169067383,
596
- 0.020039680480957032,
597
- 0.020048896789550782,
598
- 0.02008678436279297,
599
- 0.019993600845336915,
600
- 0.020132863998413086,
601
- 0.020093952178955078,
602
- 0.02027212715148926,
603
- 0.020182016372680665,
604
- 0.020141056060791016,
605
- 0.020102144241333008,
606
- 0.02011136054992676,
607
- 0.020133888244628906,
608
- 0.020090879440307616,
609
- 0.020168703079223634,
610
- 0.020102144241333008,
611
- 0.020191232681274415,
612
- 0.020114431381225584,
613
- 0.020214784622192384,
614
- 0.02020249557495117,
615
- 0.02008064079284668,
616
- 0.02005606460571289,
617
- 0.020197376251220703,
618
- 0.020033536911010744,
619
- 0.020036703109741212,
620
- 0.02002627182006836,
621
- 0.020059135437011717,
622
- 0.020083711624145507,
623
- 0.02007347106933594,
624
- 0.02012057685852051,
625
- 0.020264959335327147,
626
- 0.020137983322143553,
627
- 0.020159488677978517,
628
- 0.020551679611206054,
629
- 0.02017807960510254,
630
- 0.020067167282104493,
631
- 0.020006912231445313,
632
- 0.020017152786254884,
633
- 0.020304895401000975,
634
- 0.02004991912841797,
635
- 0.019971071243286134,
636
- 0.020060159683227538,
637
- 0.0200581111907959,
638
- 0.020142080307006836,
639
- 0.020034624099731446,
640
- 0.020093887329101563,
641
- 0.02006118392944336,
642
- 0.020176895141601564,
643
- 0.019990528106689453,
644
- 0.020172800064086914,
645
- 0.020083711624145507,
646
- 0.02006118392944336,
647
- 0.020107263565063475,
648
- 0.01998847961425781,
649
- 0.019936256408691407,
650
- 0.02008883285522461,
651
- 0.020034624099731446,
652
- 0.020571071624755858,
653
- 0.020121824264526366,
654
- 0.020138784408569334,
655
- 0.020067327499389647,
656
- 0.02006425666809082,
657
- 0.020084735870361328,
658
- 0.020116479873657226,
659
- 0.020141088485717773,
660
- 0.02015126419067383,
661
- 0.02006630325317383,
662
- 0.02007244873046875,
663
- 0.02003455924987793,
664
- 0.02004684829711914,
665
- 0.0200817928314209,
666
- 0.02024537658691406,
667
- 0.020103200912475586,
668
- 0.020157407760620118,
669
- 0.020136959075927736,
670
- 0.020133888244628906,
671
- 0.02002227210998535,
672
- 0.02007244873046875,
673
- 0.02007040023803711,
674
- 0.02007040023803711,
675
- 0.02008166313171387,
676
- 0.020060159683227538,
677
- 0.020206815719604494,
678
- 0.020085535049438476,
679
- 0.02023423957824707,
680
- 0.020200447082519533,
681
- 0.020098047256469728,
682
- 0.02004684829711914,
683
- 0.02006118392944336,
684
- 0.02007040023803711,
685
- 0.020136959075927736,
686
- 0.020144128799438478,
687
- 0.020118528366088868,
688
- 0.020288511276245116,
689
- 0.020099071502685546,
690
- 0.02019430351257324,
691
- 0.02004684829711914,
692
- 0.02005401611328125,
693
- 0.020015104293823242,
694
- 0.019991552352905274,
695
  0.019975168228149414,
696
- 0.020050943374633787,
697
- 0.020155391693115234,
698
- 0.019990528106689453,
699
- 0.02006425666809082,
700
- 0.020118528366088868,
701
- 0.020065280914306642,
702
- 0.02002227210998535,
703
- 0.019994623184204103,
704
- 0.0200447998046875,
705
- 0.019950592041015625,
706
- 0.02007756805419922,
707
- 0.019956735610961913,
708
- 0.019977216720581056,
709
- 0.020134912490844727,
710
- 0.02004684829711914,
711
- 0.020101119995117187,
712
- 0.02007347106933594,
713
- 0.020011007308959963,
714
- 0.020031488418579102,
715
- 0.020001792907714845,
716
- 0.019971071243286134,
717
- 0.019919872283935547,
718
- 0.020015104293823242,
719
- 0.019974143981933593,
720
- 0.020084735870361328,
721
  0.020006912231445313,
722
- 0.02004684829711914,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
723
  0.02002841567993164,
724
- 0.019969024658203126,
725
- 0.02004991912841797,
726
- 0.019981311798095702,
727
- 0.02044108772277832,
728
- 0.019962879180908204,
729
- 0.019952640533447266,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
730
  0.0198922233581543,
731
- 0.020084735870361328,
732
- 0.019954687118530275,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
733
  0.01999564743041992,
734
- 0.020052000045776366,
735
- 0.01995363235473633,
 
 
 
 
 
 
736
  0.01987276840209961,
 
 
 
 
 
 
 
 
737
  0.019944448471069336,
738
- 0.019919872283935547,
739
- 0.019914751052856446,
740
- 0.019826688766479493,
741
- 0.02004582405090332,
742
- 0.019941375732421874,
743
- 0.019929088592529298,
744
- 0.02001817512512207,
745
- 0.019945472717285157,
746
- 0.019958784103393554,
747
- 0.01985024070739746,
748
- 0.019937280654907227,
749
- 0.019853311538696287,
750
- 0.0199466552734375,
751
- 0.020431711196899415,
752
- 0.019922176361083985,
753
- 0.02009574317932129,
754
- 0.019914751052856446,
755
- 0.019936256408691407,
756
- 0.01992192077636719,
757
- 0.019983360290527344,
758
- 0.02001408004760742,
759
- 0.019893247604370116,
760
- 0.02000383949279785,
761
  0.01990656089782715,
762
- 0.01999977684020996,
763
- 0.019901407241821288,
764
- 0.01988198471069336,
765
  0.019887104034423828,
766
- 0.019885055541992186,
767
- 0.01984819221496582,
768
- 0.019936256408691407,
769
- 0.019959808349609375,
770
- 0.020031488418579102,
 
 
771
  0.019909631729125975,
772
- 0.02001203155517578,
773
- 0.020017152786254884,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
774
  0.01986457633972168,
775
- 0.020016128540039063,
776
- 0.01998028755187988,
777
- 0.019956735610961913,
778
- 0.019976192474365235,
779
- 0.019934207916259765,
780
- 0.019861503601074217,
781
- 0.01985740852355957,
782
- 0.02012175941467285,
783
- 0.019886943817138673,
784
- 0.019940576553344726,
785
- 0.020035360336303713,
786
  0.01983795166015625,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
787
  0.019915775299072267,
788
- 0.019947519302368166,
789
- 0.0198656005859375,
790
- 0.01988096046447754,
791
- 0.01987993621826172,
792
- 0.02002547264099121,
793
- 0.019898239135742186,
794
- 0.019997695922851562,
795
- 0.019975168228149414,
796
- 0.019931135177612306,
797
- 0.019895456314086915,
798
- 0.020028255462646485,
799
- 0.019933183670043944,
800
- 0.02004172706604004,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
801
  0.01986355209350586,
802
- 0.01991689682006836,
803
- 0.019908512115478515,
804
- 0.0198973445892334,
805
- 0.019948543548583983,
806
- 0.019912832260131835,
807
- 0.019990400314331056,
808
- 0.019948671340942383,
809
- 0.02000985527038574,
810
- 0.020107263565063475,
811
- 0.019983360290527344,
812
- 0.0200447998046875,
813
  0.019920896530151368,
814
- 0.019947519302368166,
815
- 0.019975168228149414,
816
- 0.01979916763305664,
817
- 0.020022144317626955,
818
- 0.019935232162475586,
819
- 0.02007859230041504,
820
- 0.020039680480957032,
821
- 0.019942495346069337,
822
- 0.01982374382019043,
823
- 0.0199617919921875,
824
- 0.019870559692382814,
825
- 0.01989017677307129,
826
- 0.020015104293823242,
827
- 0.02007142448425293,
828
- 0.020382720947265624,
829
- 0.02000486373901367,
830
- 0.019971071243286134,
831
- 0.019987455368041994,
832
- 0.019854335784912108,
833
- 0.01992192077636719,
834
- 0.019928064346313477,
835
- 0.019933183670043944,
836
- 0.020311040878295897,
837
- 0.019928064346313477,
838
- 0.020082687377929686,
839
- 0.02009600067138672,
840
  0.02005606460571289,
841
- 0.020016128540039063,
842
- 0.019936256408691407,
843
- 0.019959808349609375,
 
 
 
 
 
 
 
 
 
844
  0.019910655975341796,
845
- 0.019986431121826173,
846
- 0.02003455924987793,
847
- 0.02021887969970703,
848
- 0.019949567794799804,
849
- 0.0200898551940918,
850
- 0.020098047256469728,
851
- 0.02005299186706543,
852
- 0.02047385597229004,
853
- 0.02004377555847168,
854
- 0.020024320602416993,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
855
  0.020033536911010744,
856
- 0.020089887619018556,
857
- 0.020184032440185545,
858
- 0.0200614070892334,
859
- 0.02000464057922363,
860
- 0.02010316848754883,
861
- 0.02003558349609375,
862
- 0.02004275131225586,
863
- 0.020051967620849608,
864
- 0.02001420783996582,
865
- 0.020013952255249025,
866
- 0.019941375732421874,
867
- 0.020170751571655272,
868
- 0.02005299186706543,
869
- 0.020015104293823242,
870
- 0.020152320861816408,
871
- 0.02008883285522461,
872
- 0.020011007308959963,
873
- 0.02008166313171387,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
874
  0.0200447998046875,
875
- 0.020098047256469728,
 
 
 
 
 
 
 
 
 
 
 
876
  0.02007859230041504,
877
- 0.020091903686523437,
878
- 0.020031488418579102,
879
- 0.02004377555847168,
880
- 0.0200581111907959,
881
- 0.020031488418579102,
882
- 0.020110336303710938,
883
  0.019987455368041994,
884
- 0.020008096694946288,
885
- 0.02011734390258789,
886
- 0.020011007308959963,
887
- 0.020099071502685546,
888
- 0.020084735870361328,
889
  0.019981311798095702,
890
- 0.020151296615600587,
891
- 0.019987455368041994,
892
- 0.020060159683227538,
893
- 0.020139007568359374,
894
- 0.020107263565063475,
895
- 0.02007347106933594,
896
- 0.0200130558013916,
897
- 0.01998847961425781,
898
- 0.020039680480957032,
899
- 0.02003558349609375,
900
- 0.020207616806030275,
901
- 0.019977216720581056,
902
- 0.020002815246582033,
903
  0.019982336044311523,
904
- 0.020069599151611328,
905
- 0.019963680267333986,
906
- 0.02016979217529297,
907
- 0.01995871925354004,
908
- 0.02005504035949707,
909
- 0.020121599197387697,
910
- 0.02000899124145508,
911
- 0.019979232788085936,
912
- 0.020040704727172853,
913
- 0.020023296356201172,
914
- 0.019956735610961913,
915
- 0.02002124786376953,
916
- 0.020188159942626953,
917
- 0.019956735610961913,
918
- 0.02007244873046875,
919
- 0.020107263565063475,
920
- 0.02023324775695801,
921
- 0.020006879806518555,
922
- 0.020163583755493163,
923
- 0.020126720428466797,
924
- 0.020001792907714845,
925
- 0.0200130558013916,
926
- 0.02001408004760742,
927
  0.019965951919555663,
928
- 0.020163583755493163,
929
- 0.020410367965698242,
930
- 0.020028608322143555,
931
- 0.020106048583984376,
932
- 0.02002841567993164,
933
- 0.01998028755187988,
934
- 0.01998355293273926,
935
- 0.020022079467773436,
936
- 0.019979263305664064,
937
- 0.020106239318847655,
938
- 0.02002227210998535,
939
- 0.020009983062744142,
940
- 0.020166656494140626,
941
- 0.020040704727172853,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
942
  0.02002841567993164,
943
- 0.02011238479614258,
944
- 0.020114431381225584,
945
- 0.02003865623474121,
 
 
946
  0.020017152786254884,
947
- 0.020000768661499024,
 
 
 
 
 
 
 
948
  0.019999744415283204,
949
- 0.02009600067138672,
950
- 0.02005504035949707,
951
- 0.019989503860473632,
952
- 0.020221952438354493,
953
- 0.020342784881591795
 
 
954
  ]
955
  },
956
  "throughput": {
957
  "unit": "tokens/s",
958
- "value": 49.87662463780578
959
  },
960
  "energy": null,
961
  "efficiency": null
 
2
  "prefill": {
3
  "memory": {
4
  "unit": "MB",
5
+ "max_ram": 1079.64416,
6
+ "max_vram": 2799.69792,
7
  "max_reserved": 2709.520384,
8
  "max_allocated": 2546.903552
9
  },
10
  "latency": {
11
  "unit": "s",
12
+ "mean": 0.02572149676835321,
13
+ "stdev": 0.0003832267571204168,
14
  "values": [
15
+ 0.028890047073364258,
16
+ 0.02668339157104492,
17
+ 0.02630143928527832,
18
+ 0.02615910339355469,
19
+ 0.026649599075317384,
20
+ 0.026203136444091796,
21
+ 0.026236928939819337,
22
+ 0.026282079696655275,
23
+ 0.026228736877441407,
24
+ 0.02814668846130371,
25
+ 0.02917580795288086,
26
+ 0.027281408309936524,
27
+ 0.026927104949951174,
28
+ 0.026056703567504884,
29
+ 0.025590784072875978,
30
+ 0.025614336013793947,
31
+ 0.025592832565307616,
32
+ 0.025585664749145507,
33
+ 0.025563135147094726,
34
+ 0.025585664749145507,
35
+ 0.025606271743774413,
36
+ 0.025584640502929686,
37
+ 0.0255346565246582,
38
+ 0.02552934455871582,
39
+ 0.025576448440551756,
40
+ 0.025665536880493164,
41
+ 0.025590784072875978,
42
+ 0.02556723213195801,
43
+ 0.025762815475463868,
44
+ 0.02734489631652832,
45
+ 0.02728447914123535,
46
+ 0.027280384063720704,
47
+ 0.027611135482788086,
48
+ 0.02730803108215332,
49
+ 0.027356159210205077,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
50
  0.025646080017089845,
51
+ 0.025540607452392578,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
  0.025648128509521483,
 
 
53
  0.025654272079467775,
54
  0.025622528076171876,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
55
  0.025659391403198242,
56
+ 0.02556620788574219,
57
+ 0.025624767303466797,
58
+ 0.02564121627807617,
59
+ 0.025820159912109376,
60
+ 0.02573721694946289,
61
+ 0.02579052734375,
62
+ 0.025836544036865236,
63
+ 0.02574028778076172,
64
+ 0.025762815475463868,
65
+ 0.02573107147216797,
66
+ 0.02573209571838379,
67
+ 0.02572902488708496,
68
+ 0.02569932746887207,
69
+ 0.02569625663757324,
70
+ 0.025778175354003906,
71
  0.02571468734741211,
72
+ 0.02568704032897949,
73
+ 0.02578041648864746,
74
+ 0.025785343170166015,
75
+ 0.02572902488708496,
76
+ 0.025745407104492187,
 
 
 
 
 
 
 
77
  0.0257126407623291,
78
+ 0.026137599945068358,
79
+ 0.025745407104492187,
80
  0.025683967590332032,
81
+ 0.02571468734741211,
82
+ 0.02568704032897949,
83
+ 0.02572083282470703,
84
+ 0.02572390365600586,
85
+ 0.02569625663757324,
86
+ 0.02570649528503418,
87
+ 0.025680896759033203,
88
  0.025686016082763673,
89
+ 0.025705472946166992,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
90
  0.02569523239135742,
91
+ 0.0257108154296875,
92
+ 0.025651199340820312,
93
  0.02574131202697754,
94
+ 0.025701375961303712,
95
+ 0.02573619270324707,
96
+ 0.025683967590332032,
97
+ 0.02608639907836914,
 
 
 
 
 
 
 
 
 
 
98
  0.02574131202697754,
99
+ 0.025614336013793947,
100
+ 0.02576383972167969,
101
+ 0.025607168197631838,
102
+ 0.02572902488708496,
103
+ 0.025597951889038087,
104
+ 0.025711679458618165,
105
+ 0.025771007537841797,
106
+ 0.025775104522705077,
107
+ 0.02570240020751953,
108
+ 0.02569932746887207,
109
  0.025667583465576172,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
110
  0.025667583465576172,
111
+ 0.02569523239135742,
112
+ 0.02575257682800293,
113
+ 0.02575155258178711,
114
+ 0.025738239288330078,
115
+ 0.025787391662597657,
116
+ 0.02574336051940918,
117
+ 0.025635839462280274,
118
+ 0.02570969581604004,
119
+ 0.025753599166870117,
120
+ 0.02574131202697754,
121
+ 0.025766912460327147,
122
+ 0.025802751541137696,
123
+ 0.025802751541137696,
124
+ 0.025746431350708008,
125
+ 0.025797632217407225,
126
+ 0.0257392635345459,
127
+ 0.025790464401245116,
128
+ 0.02577305603027344,
129
+ 0.025831424713134765,
130
+ 0.02575564765930176,
131
+ 0.025785343170166015,
132
+ 0.025781343460083008,
133
+ 0.02573209571838379,
134
+ 0.02569932746887207,
135
+ 0.025661632537841796,
136
+ 0.025701375961303712,
137
+ 0.025853952407836913,
138
+ 0.025734144210815428,
139
+ 0.02576076889038086,
140
+ 0.025786367416381836,
141
+ 0.025676799774169923,
142
+ 0.02567987251281738,
143
+ 0.025665536880493164,
144
+ 0.02576896095275879,
145
+ 0.02567884826660156,
146
+ 0.02571571159362793,
147
+ 0.025628671646118165,
148
+ 0.025622528076171876,
149
+ 0.025651199340820312,
150
+ 0.02569011116027832,
151
+ 0.025640960693359374,
152
+ 0.02565836715698242,
153
+ 0.025792512893676758,
154
+ 0.02570035171508789,
155
  0.025649152755737304,
156
+ 0.02614169692993164,
 
 
 
 
 
157
  0.025643007278442383,
158
+ 0.02571571159362793,
159
+ 0.025676799774169923,
160
+ 0.02572902488708496,
161
+ 0.025694208145141603,
162
+ 0.02566655921936035,
163
+ 0.02570444869995117,
164
+ 0.025782272338867186,
165
+ 0.02568294334411621,
166
+ 0.02568704032897949,
167
+ 0.025669631958007814,
168
+ 0.025652223587036133,
169
+ 0.025639936447143553,
170
+ 0.025573375701904297,
171
  0.025622528076171876,
172
+ 0.025592832565307616,
173
+ 0.025647104263305662,
174
+ 0.025582592010498048,
175
+ 0.026016767501831056,
176
+ 0.025745407104492187,
177
+ 0.025665536880493164,
178
+ 0.025677824020385744,
179
+ 0.025671680450439452,
180
+ 0.02553753662109375,
181
+ 0.02568704032897949,
182
+ 0.02555904006958008,
183
+ 0.025673728942871094,
184
+ 0.025573375701904297,
185
+ 0.02550579261779785,
186
+ 0.025601024627685546,
187
+ 0.025628671646118165,
188
+ 0.02553241539001465,
189
+ 0.02549452781677246,
190
+ 0.025554943084716796,
191
+ 0.025570304870605468,
192
  0.025585664749145507,
193
+ 0.02576896095275879,
194
+ 0.025539743423461915,
195
+ 0.025636863708496094,
196
+ 0.025622528076171876,
197
+ 0.025600000381469725,
198
+ 0.025604095458984375,
199
+ 0.025600000381469725,
200
+ 0.02555904006958008,
201
+ 0.025622560501098634,
202
+ 0.02554368019104004,
203
+ 0.025625600814819335,
204
+ 0.025581567764282227,
205
+ 0.025632768630981444,
206
+ 0.025593856811523437,
207
+ 0.025608192443847655,
208
+ 0.02557439994812012,
209
+ 0.025671680450439452,
210
+ 0.02554572868347168,
211
+ 0.025630720138549806,
212
+ 0.025611263275146484,
213
+ 0.025653247833251954,
214
+ 0.02558361625671387,
215
+ 0.025612287521362305,
216
+ 0.025631744384765624,
217
+ 0.025652223587036133,
218
+ 0.025574464797973633,
219
  0.025624576568603515,
220
+ 0.025563135147094726,
221
+ 0.025669631958007814,
222
+ 0.025625600814819335,
223
+ 0.025630943298339842,
224
+ 0.025596927642822266,
225
  0.025601024627685546,
226
+ 0.025588735580444336,
227
+ 0.025612287521362305,
228
+ 0.025616384506225585,
229
+ 0.025608192443847655,
230
+ 0.025597951889038087,
231
+ 0.025539583206176757,
232
+ 0.025601024627685546,
233
+ 0.025565183639526368,
234
  0.025631744384765624,
235
+ 0.02595123291015625,
236
+ 0.02571161651611328,
237
+ 0.02555187225341797,
238
+ 0.025544704437255858,
239
+ 0.025500831604003907,
240
+ 0.025621503829956056,
241
+ 0.02555904006958008,
242
+ 0.025540607452392578,
243
+ 0.025596063613891603,
244
+ 0.025582592010498048,
245
+ 0.026077184677124023,
246
+ 0.025655296325683592,
247
+ 0.025597951889038087,
248
+ 0.02569932746887207,
249
+ 0.025513055801391602,
250
+ 0.025503744125366212,
251
+ 0.025531391143798828,
252
  0.025571327209472656,
253
+ 0.025578496932983398,
254
+ 0.02597068786621094,
 
 
 
 
 
 
 
 
255
  0.025648128509521483,
 
256
  0.025645055770874024,
257
+ 0.02566771125793457,
 
 
 
 
 
258
  0.025585664749145507,
259
+ 0.025573375701904297,
260
+ 0.02555392074584961,
261
+ 0.02552524757385254,
262
+ 0.025621503829956056,
263
+ 0.025613311767578126,
264
+ 0.025616384506225585,
265
+ 0.02556620788574219,
266
+ 0.025585664749145507,
267
+ 0.025588735580444336,
268
+ 0.025597951889038087,
269
+ 0.025491455078125,
270
  0.025587711334228515,
271
+ 0.025569280624389647,
272
+ 0.025621696472167967,
273
+ 0.02553446388244629,
274
+ 0.025556991577148438,
275
+ 0.025585664749145507,
276
+ 0.025651199340820312,
277
+ 0.02555801582336426,
278
+ 0.025611263275146484,
279
+ 0.025591808319091795,
280
+ 0.025594879150390625,
281
+ 0.025648128509521483,
282
+ 0.025689088821411132,
283
+ 0.02556620788574219,
284
+ 0.025648128509521483,
285
  0.025635839462280274,
286
+ 0.025634815216064453,
287
+ 0.02552217674255371,
288
+ 0.02553036880493164,
289
+ 0.025564159393310547,
290
+ 0.025633792877197265,
291
+ 0.02569625663757324,
292
+ 0.025635839462280274,
293
+ 0.025643007278442383,
294
+ 0.025656415939331056,
295
+ 0.025568256378173827,
 
 
 
 
 
 
 
 
 
 
 
 
 
296
  0.025629695892333985,
297
+ 0.025589759826660157,
298
+ 0.025684991836547853,
299
+ 0.025614336013793947,
300
+ 0.025576448440551756,
301
+ 0.02552115249633789,
302
+ 0.02557439994812012,
303
+ 0.025592832565307616,
304
+ 0.02574847984313965,
305
+ 0.02559715270996094,
306
+ 0.025842687606811524,
307
+ 0.02555801582336426,
308
+ 0.02553548812866211,
309
+ 0.025587711334228515,
310
+ 0.02570240020751953,
311
+ 0.025585664749145507,
312
+ 0.02557542419433594,
313
+ 0.025528383255004884,
314
+ 0.025919679641723634,
315
+ 0.025657344818115234,
316
+ 0.025657567977905273,
317
+ 0.025588735580444336,
318
+ 0.025585664749145507,
319
+ 0.025560064315795897,
320
+ 0.025555967330932617,
321
+ 0.025631744384765624,
322
+ 0.02567884826660156,
323
+ 0.025589759826660157,
324
+ 0.025636863708496094,
325
+ 0.025606143951416017,
326
+ 0.02554982376098633,
327
+ 0.0255467529296875,
328
  0.025552896499633788,
329
+ 0.025598976135253908,
330
+ 0.02567884826660156,
331
+ 0.025630720138549806,
332
+ 0.025975807189941406,
333
+ 0.025577472686767577,
334
+ 0.025572351455688477,
335
+ 0.025571327209472656,
336
+ 0.025638912200927736,
337
+ 0.025658464431762694,
338
+ 0.025573375701904297,
339
+ 0.025561088562011718,
340
+ 0.025565183639526368,
341
+ 0.025593856811523437,
342
+ 0.025609216690063476,
343
+ 0.025555967330932617,
344
+ 0.02557439994812012,
345
+ 0.02553049659729004,
346
+ 0.02551603126525879,
347
+ 0.02556211280822754,
348
  0.025625600814819335,
349
+ 0.02554572868347168,
350
+ 0.025625600814819335,
351
+ 0.025531391143798828,
352
+ 0.025605119705200196,
353
+ 0.025553951263427733,
354
+ 0.025651199340820312,
355
+ 0.025586687088012695,
356
+ 0.025573375701904297,
357
+ 0.025586687088012695,
358
+ 0.025564159393310547,
359
+ 0.02555801582336426,
360
+ 0.025652223587036133,
361
+ 0.0255283203125,
362
+ 0.025648128509521483,
363
+ 0.025615360260009764,
364
+ 0.025589759826660157,
365
+ 0.025584640502929686,
366
+ 0.025693183898925782,
367
+ 0.025564159393310547,
368
  0.02570854377746582,
369
+ 0.025634815216064453,
370
+ 0.025624576568603515,
371
+ 0.02550579261779785,
372
+ 0.0255098876953125,
373
+ 0.025552896499633788,
374
+ 0.025606143951416017,
375
+ 0.025653247833251954,
376
+ 0.025580543518066406,
377
+ 0.025627647399902344,
378
  0.025614336013793947,
379
+ 0.025554943084716796,
380
+ 0.025602048873901367,
381
+ 0.02558380889892578,
382
  0.0257259521484375,
383
+ 0.025611263275146484,
384
+ 0.02557542419433594,
385
+ 0.025597951889038087,
386
+ 0.025581567764282227,
387
+ 0.025553056716918945,
388
+ 0.025686016082763673,
389
+ 0.025618431091308593,
390
+ 0.025630847930908203,
391
+ 0.02554982376098633,
392
+ 0.026102783203125,
393
+ 0.025620479583740235,
394
+ 0.025677824020385744,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
395
  0.025660415649414063,
396
+ 0.025615360260009764,
397
+ 0.025628671646118165,
398
+ 0.025592832565307616,
399
+ 0.025610240936279297,
400
+ 0.02572185516357422,
401
+ 0.025588735580444336,
402
+ 0.025628671646118165,
403
+ 0.025577600479125977
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
404
  ]
405
  },
406
  "throughput": {
407
  "unit": "tokens/s",
408
+ "value": 9952.764502996304
409
  },
410
  "energy": null,
411
  "efficiency": null
 
413
  "decode": {
414
  "memory": {
415
  "unit": "MB",
416
+ "max_ram": 1079.64416,
417
  "max_vram": 3472.883712,
418
  "max_reserved": 2988.4416,
419
  "max_allocated": 2813.773312
420
  },
421
  "latency": {
422
  "unit": "s",
423
+ "mean": 5.0872309780120855,
424
+ "stdev": 0.0026828880310065806,
425
  "values": [
426
+ 5.084548089981078,
427
+ 5.089913866043092
428
  ]
429
  },
430
  "throughput": {
431
  "unit": "tokens/s",
432
+ "value": 50.1255007099452
433
  },
434
  "energy": null,
435
  "efficiency": null
436
+ },
437
+ "per_token": {
438
+ "memory": null,
439
+ "latency": {
440
+ "unit": "s",
441
+ "mean": 0.019949925403968952,
442
+ "stdev": 0.00014777607341213587,
443
+ "values": [
444
+ 0.02022809600830078,
445
+ 0.020031776428222656,
446
+ 0.019912416458129883,
447
+ 0.019926111221313478,
448
+ 0.019920799255371095,
449
+ 0.019964927673339843,
450
+ 0.019896320343017578,
451
+ 0.019878911972045898,
452
+ 0.020257823944091796,
453
+ 0.019924959182739257,
454
+ 0.02003660774230957,
455
+ 0.01987174415588379,
456
+ 0.019905664443969726,
457
+ 0.019874687194824218,
458
+ 0.019901439666748046,
459
+ 0.019911872863769532,
460
+ 0.019849023818969726,
461
+ 0.019944448471069336,
462
+ 0.01986355209350586,
463
+ 0.019960895538330078,
464
+ 0.01995564842224121,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
465
  0.019975168228149414,
466
+ 0.019959808349609375,
467
+ 0.019966976165771484,
468
+ 0.020130815505981444,
469
+ 0.020016128540039063,
470
+ 0.020008031845092773,
471
+ 0.01999555206298828,
472
+ 0.019962879180908204,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
473
  0.020006912231445313,
474
+ 0.020039680480957032,
475
+ 0.019907583236694337,
476
+ 0.019886240005493164,
477
+ 0.019990367889404295,
478
+ 0.019982336044311523,
479
+ 0.019989503860473632,
480
+ 0.019862527847290038,
481
+ 0.019900415420532228,
482
+ 0.019900415420532228,
483
+ 0.01998863983154297,
484
+ 0.019859296798706055,
485
+ 0.019939327239990236,
486
+ 0.019917823791503905,
487
+ 0.020121599197387697,
488
+ 0.019936256408691407,
489
+ 0.020023296356201172,
490
+ 0.019986560821533203,
491
+ 0.019899263381958007,
492
+ 0.019937280654907227,
493
+ 0.019911680221557617,
494
+ 0.019893247604370116,
495
+ 0.019961856842041017,
496
+ 0.019919872283935547,
497
+ 0.019940351486206053,
498
  0.02002841567993164,
499
+ 0.020059135437011717,
500
+ 0.019932159423828123,
501
+ 0.019909631729125975,
502
+ 0.02004595184326172,
503
+ 0.01989414405822754,
504
+ 0.019945472717285157,
505
+ 0.019904544830322266,
506
+ 0.019898336410522462,
507
+ 0.019940351486206053,
508
+ 0.02002227210998535,
509
+ 0.019948543548583983,
510
+ 0.019976192474365235,
511
+ 0.019911840438842775,
512
+ 0.019958623886108397,
513
+ 0.02009712028503418,
514
+ 0.01982352066040039,
515
+ 0.019912704467773438,
516
+ 0.01987481689453125,
517
+ 0.01988115119934082,
518
+ 0.019920703887939453,
519
+ 0.019974143981933593,
520
  0.0198922233581543,
521
+ 0.019901439666748046,
522
+ 0.019942399978637695,
523
+ 0.019895328521728515,
524
+ 0.01992291259765625,
525
+ 0.01982076835632324,
526
+ 0.019967775344848632,
527
+ 0.019910655975341796,
528
+ 0.019999744415283204,
529
+ 0.02047305679321289,
530
+ 0.02004047966003418,
531
+ 0.019919071197509765,
532
+ 0.019970848083496095,
533
+ 0.0199979190826416,
534
+ 0.01995564842224121,
535
+ 0.020112415313720704,
536
+ 0.019997503280639647,
537
+ 0.019976192474365235,
538
+ 0.02007859230041504,
539
+ 0.019928064346313477,
540
+ 0.019971168518066407,
541
+ 0.019905439376831056,
542
+ 0.01982771110534668,
543
+ 0.01986355209350586,
544
+ 0.0198922233581543,
545
+ 0.019911680221557617,
546
+ 0.019876863479614256,
547
+ 0.019885055541992186,
548
+ 0.019843072891235353,
549
  0.01999564743041992,
550
+ 0.0198656005859375,
551
+ 0.01987583923339844,
552
+ 0.019923967361450197,
553
+ 0.02028544044494629,
554
+ 0.01985843276977539,
555
+ 0.019934207916259765,
556
+ 0.019914751052856446,
557
+ 0.019877887725830077,
558
  0.01987276840209961,
559
+ 0.019916799545288084,
560
+ 0.019962015151977538,
561
+ 0.019856224060058592,
562
+ 0.01988915252685547,
563
+ 0.019923967361450197,
564
+ 0.019959808349609375,
565
+ 0.019908607482910155,
566
+ 0.019876863479614256,
567
  0.019944448471069336,
568
+ 0.019976192474365235,
569
+ 0.019984384536743165,
570
+ 0.01998355293273926,
571
+ 0.020041568756103516,
572
+ 0.01992188835144043,
573
+ 0.019928064346313477,
574
+ 0.019923967361450197,
575
+ 0.01990355110168457,
576
+ 0.01991878318786621,
577
+ 0.019900415420532228,
578
+ 0.01987276840209961,
579
+ 0.019923967361450197,
 
 
 
 
 
 
 
 
 
 
 
580
  0.01990656089782715,
581
+ 0.01998847961425781,
 
 
582
  0.019887104034423828,
583
+ 0.019895296096801757,
584
+ 0.01979084777832031,
585
+ 0.019939327239990236,
586
+ 0.019907583236694337,
587
+ 0.019917823791503905,
588
+ 0.019876960754394532,
589
+ 0.01984604835510254,
590
  0.019909631729125975,
591
+ 0.019932159423828123,
592
+ 0.019808256149291992,
593
+ 0.019927040100097656,
594
+ 0.019796159744262694,
595
+ 0.0197640323638916,
596
+ 0.01982156753540039,
597
+ 0.01982975959777832,
598
+ 0.019786752700805665,
599
+ 0.019834880828857423,
600
+ 0.01987174415588379,
601
+ 0.01986867141723633,
602
+ 0.01993011283874512,
603
+ 0.019860607147216797,
604
+ 0.019815296173095704,
605
+ 0.01984332847595215,
606
+ 0.019841791152954102,
607
+ 0.019772415161132813,
608
+ 0.01982771110534668,
609
+ 0.01987379264831543,
610
+ 0.019979263305664064,
611
+ 0.019910655975341796,
612
+ 0.019853311538696287,
613
+ 0.019787776947021486,
614
+ 0.019892383575439453,
615
+ 0.019857248306274413,
616
+ 0.01984102439880371,
617
+ 0.019997695922851562,
618
+ 0.019973119735717772,
619
+ 0.019989503860473632,
620
+ 0.019854335784912108,
621
+ 0.019878911972045898,
622
+ 0.019818496704101563,
623
+ 0.019901439666748046,
624
+ 0.02006425666809082,
625
+ 0.019885055541992186,
626
+ 0.01997532844543457,
627
+ 0.01992073631286621,
628
+ 0.02000383949279785,
629
+ 0.019883007049560548,
630
+ 0.020426816940307617,
631
+ 0.01997305679321289,
632
+ 0.019902463912963866,
633
+ 0.019893247604370116,
634
+ 0.019918848037719726,
635
+ 0.019883007049560548,
636
+ 0.019808256149291992,
637
+ 0.01984422492980957,
638
+ 0.0200948486328125,
639
+ 0.01983078384399414,
640
  0.01986457633972168,
641
+ 0.019817472457885742,
642
+ 0.019911680221557617,
643
+ 0.01988403129577637,
644
+ 0.019869695663452147,
645
+ 0.020044960021972657,
646
+ 0.019830623626708985,
647
+ 0.019962879180908204,
648
+ 0.019903488159179687,
 
 
 
649
  0.01983795166015625,
650
+ 0.019973119735717772,
651
+ 0.019900415420532228,
652
+ 0.019954687118530275,
653
+ 0.01982464027404785,
654
+ 0.02046156883239746,
655
+ 0.019920896530151368,
656
+ 0.019912704467773438,
657
+ 0.01991372871398926,
658
+ 0.019910655975341796,
659
+ 0.019941375732421874,
660
+ 0.01998041534423828,
661
+ 0.019852319717407228,
662
+ 0.020055072784423828,
663
+ 0.01996086311340332,
664
+ 0.019908544540405273,
665
+ 0.01996272087097168,
666
  0.019915775299072267,
667
+ 0.0200130558013916,
668
+ 0.01990060806274414,
669
+ 0.019860288619995118,
670
+ 0.01983283233642578,
671
+ 0.019963903427124022,
672
+ 0.019884096145629884,
673
+ 0.01998841667175293,
674
+ 0.020002944946289063,
675
+ 0.02009280014038086,
676
+ 0.019969024658203126,
677
+ 0.020059135437011717,
678
+ 0.019937280654907227,
679
+ 0.020118528366088868,
680
+ 0.02010316848754883,
681
+ 0.0200898551940918,
682
+ 0.020007007598876952,
683
+ 0.01988921546936035,
684
+ 0.019886943817138673,
685
+ 0.020001792907714845,
686
+ 0.020002944946289063,
687
+ 0.019973215103149415,
688
+ 0.020085695266723633,
689
+ 0.020010047912597657,
690
+ 0.02006831932067871,
691
+ 0.019857248306274413,
692
+ 0.01991891288757324,
693
+ 0.019940319061279296,
694
+ 0.01993824005126953,
695
+ 0.019941375732421874,
696
+ 0.019936256408691407,
697
+ 0.020122623443603514,
698
+ 0.019963903427124022,
699
+ 0.020117664337158205,
700
+ 0.01995657539367676,
701
+ 0.019965951919555663,
702
+ 0.019960832595825196,
703
+ 0.019950592041015625,
704
+ 0.019996896743774414,
705
+ 0.01985728073120117,
706
+ 0.01980918312072754,
707
+ 0.01984511947631836,
708
+ 0.01990982437133789,
709
+ 0.019873600006103515,
710
+ 0.019911903381347656,
711
+ 0.019950368881225585,
712
+ 0.01996294403076172,
713
+ 0.019907520294189452,
714
+ 0.01983897590637207,
715
+ 0.01984921646118164,
716
+ 0.01988812828063965,
717
+ 0.01990553665161133,
718
+ 0.019992576599121094,
719
  0.01986355209350586,
720
+ 0.0198604793548584,
721
+ 0.019890207290649414,
722
+ 0.019909599304199218,
723
+ 0.01988403129577637,
724
+ 0.01986764717102051,
725
+ 0.02001513671875,
726
+ 0.019925983428955078,
727
+ 0.02003660774230957,
728
+ 0.019960832595825196,
 
 
729
  0.019920896530151368,
730
+ 0.020509695053100584,
731
+ 0.019904512405395508,
732
+ 0.019929088592529298,
733
+ 0.01988915252685547,
734
+ 0.019914751052856446,
735
+ 0.019895328521728515,
736
+ 0.019859424591064455,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
737
  0.02005606460571289,
738
+ 0.019943424224853516,
739
+ 0.02000486373901367,
740
+ 0.019943424224853516,
741
+ 0.019929088592529298,
742
+ 0.0198604793548584,
743
+ 0.019891424179077147,
744
+ 0.019995424270629884,
745
+ 0.019938304901123048,
746
+ 0.019878911972045898,
747
+ 0.019813375473022463,
748
+ 0.01983897590637207,
749
+ 0.019955711364746095,
750
  0.019910655975341796,
751
+ 0.019991552352905274,
752
+ 0.01987481689453125,
753
+ 0.019910655975341796,
754
+ 0.020447231292724608,
755
+ 0.020009983062744142,
756
+ 0.01987491226196289,
757
+ 0.01983273506164551,
758
+ 0.0198922233581543,
759
+ 0.019927040100097656,
760
+ 0.019883007049560548,
761
+ 0.019781631469726564,
762
+ 0.019992576599121094,
763
+ 0.019894271850585937,
764
+ 0.01987811279296875,
765
+ 0.01980291175842285,
766
+ 0.01984409523010254,
767
+ 0.019942527770996095,
768
+ 0.0198654727935791,
769
+ 0.02025369644165039,
770
+ 0.02229964828491211,
771
+ 0.020340736389160157,
772
+ 0.01983692741394043,
773
+ 0.019971071243286134,
774
+ 0.01986662483215332,
775
+ 0.019933408737182617,
776
+ 0.019889951705932617,
777
+ 0.019870975494384765,
778
+ 0.01995955276489258,
779
  0.020033536911010744,
780
+ 0.019781631469726564,
781
+ 0.019909856796264648,
782
+ 0.019749664306640626,
783
+ 0.019715072631835938,
784
+ 0.019910655975341796,
785
+ 0.020050943374633787,
786
+ 0.01986764717102051,
787
+ 0.01991372871398926,
788
+ 0.01985024070739746,
789
+ 0.01989017677307129,
790
+ 0.0198154239654541,
791
+ 0.019894271850585937,
792
+ 0.01988403129577637,
793
+ 0.019818719863891603,
794
+ 0.019869472503662108,
795
+ 0.019920896530151368,
796
+ 0.019984384536743165,
797
+ 0.019803136825561524,
798
+ 0.019932159423828123,
799
+ 0.01992812728881836,
800
+ 0.02003468894958496,
801
+ 0.019946304321289063,
802
+ 0.01988915252685547,
803
+ 0.019773439407348634,
804
+ 0.019833856582641602,
805
+ 0.01990656089782715,
806
+ 0.01984739112854004,
807
+ 0.020088607788085938,
808
+ 0.019904512405395508,
809
+ 0.019961856842041017,
810
+ 0.019902463912963866,
811
+ 0.01993011283874512,
812
+ 0.02005606460571289,
813
+ 0.019934207916259765,
814
+ 0.02006630325317383,
815
+ 0.019932416915893553,
816
+ 0.019945215225219727,
817
+ 0.019870719909667968,
818
+ 0.0198604793548584,
819
+ 0.01987993621826172,
820
+ 0.019793920516967774,
821
+ 0.01983897590637207,
822
+ 0.01988915252685547,
823
+ 0.01992192077636719,
824
+ 0.01985024070739746,
825
+ 0.01985024070739746,
826
+ 0.01986355209350586,
827
+ 0.01999667167663574,
828
+ 0.01985740852355957,
829
+ 0.020186208724975587,
830
+ 0.019893247604370116,
831
+ 0.020323232650756837,
832
+ 0.02008086395263672,
833
+ 0.019963680267333986,
834
+ 0.0198656005859375,
835
+ 0.01987583923339844,
836
+ 0.019883007049560548,
837
+ 0.019893247604370116,
838
+ 0.019981311798095702,
839
+ 0.020021472930908203,
840
+ 0.02008963203430176,
841
+ 0.01997209548950195,
842
+ 0.020205568313598633,
843
+ 0.019990528106689453,
844
  0.0200447998046875,
845
+ 0.019951616287231445,
846
+ 0.019912704467773438,
847
+ 0.01983180809020996,
848
+ 0.019896320343017578,
849
+ 0.019903488159179687,
850
+ 0.019936256408691407,
851
+ 0.019869695663452147,
852
+ 0.02007244873046875,
853
+ 0.019918048858642578,
854
+ 0.019906335830688477,
855
+ 0.020388864517211915,
856
+ 0.020060159683227538,
857
  0.02007859230041504,
 
 
 
 
 
 
858
  0.019987455368041994,
859
+ 0.019978399276733397,
860
+ 0.019898208618164062,
861
+ 0.02007859230041504,
862
+ 0.019900415420532228,
863
+ 0.020033536911010744,
864
  0.019981311798095702,
865
+ 0.01985228729248047,
866
+ 0.019795967102050782,
867
+ 0.01988915252685547,
868
+ 0.01986764717102051,
869
+ 0.019931360244750975,
870
+ 0.019887903213500976,
871
+ 0.019968000411987305,
872
+ 0.01983283233642578,
873
+ 0.019870943069458007,
874
+ 0.02007427215576172,
875
+ 0.019926015853881835,
 
 
876
  0.019982336044311523,
877
+ 0.019936256408691407,
878
+ 0.019976192474365235,
879
+ 0.019958784103393554,
880
+ 0.01996700859069824,
881
+ 0.01986761665344238,
882
+ 0.019877887725830077,
883
+ 0.02000486373901367,
884
+ 0.019909631729125975,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
885
  0.019965951919555663,
886
+ 0.019869695663452147,
887
+ 0.020110336303710938,
888
+ 0.02000486373901367,
889
+ 0.019959808349609375,
890
+ 0.020015104293823242,
891
+ 0.019963903427124022,
892
+ 0.02002124786376953,
893
+ 0.019910655975341796,
894
+ 0.01991379165649414,
895
+ 0.019892160415649413,
896
+ 0.019946495056152345,
897
+ 0.020167680740356447,
898
+ 0.020032512664794923,
899
+ 0.01997209548950195,
900
+ 0.019894271850585937,
901
+ 0.019977216720581056,
902
+ 0.019938304901123048,
903
+ 0.019885055541992186,
904
+ 0.019870719909667968,
905
+ 0.01990163230895996,
906
+ 0.019895103454589842,
907
+ 0.02008064079284668,
908
+ 0.01987379264831543,
909
+ 0.02005833625793457,
910
+ 0.019920671463012695,
911
+ 0.0200949764251709,
912
+ 0.020108448028564454,
913
+ 0.020037471771240236,
914
+ 0.020031488418579102,
915
+ 0.019954687118530275,
916
+ 0.019946495056152345,
917
+ 0.019982336044311523,
918
+ 0.020050943374633787,
919
+ 0.01987174415588379,
920
+ 0.019982336044311523,
921
+ 0.019999744415283204,
922
+ 0.020000768661499024,
923
+ 0.020060159683227538,
924
+ 0.019922943115234376,
925
+ 0.019964927673339843,
926
+ 0.020002815246582033,
927
+ 0.020160512924194338,
928
+ 0.019947744369506835,
929
+ 0.01989200019836426,
930
+ 0.01988915252685547,
931
  0.02002841567993164,
932
+ 0.020544511795043945,
933
+ 0.019938304901123048,
934
+ 0.01998847961425781,
935
+ 0.019936256408691407,
936
+ 0.019950592041015625,
937
  0.020017152786254884,
938
+ 0.019947519302368166,
939
+ 0.019971103668212892,
940
+ 0.019946464538574217,
941
+ 0.019982336044311523,
942
+ 0.020093952178955078,
943
+ 0.019943424224853516,
944
+ 0.0200581111907959,
945
+ 0.019957759857177734,
946
  0.019999744415283204,
947
+ 0.02004172706604004,
948
+ 0.0200130558013916,
949
+ 0.020011007308959963,
950
+ 0.020140031814575195,
951
+ 0.019901439666748046,
952
+ 0.02004172706604004,
953
+ 0.02001919937133789
954
  ]
955
  },
956
  "throughput": {
957
  "unit": "tokens/s",
958
+ "value": 50.12550070994523
959
  },
960
  "energy": null,
961
  "efficiency": null