-
Notifications
You must be signed in to change notification settings - Fork 0
/
benchmark_temperature_13.txt
1769 lines (1697 loc) · 75.7 KB
/
benchmark_temperature_13.txt
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
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
Open ttft file
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.5526427030563354
P50 ttft = 0.45134222507476807
P99 ttft = 1.0483049845695496
Average tbt = 0.14899750947952276
P50 tbt = 0.020221960544586182
P99 tbt = 0.6059496738910678
All GPUs and memories are cold after 4.044384956359863
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.6575875963483538
P50 ttft = 0.5172936916351318
P99 ttft = 1.3662791252136233
Average tbt = 0.3544172548112416
P50 tbt = 0.3269990205764771
P99 tbt = 1.1903758859634403
All GPUs and memories are cold after 18.118577480316162
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.9568144934517997
P50 ttft = 0.7913851737976074
P99 ttft = 1.9860135841369617
Average tbt = 0.938033633232117
P50 tbt = 0.6338698863983157
P99 tbt = 3.6218932938575743
All GPUs and memories are cold after 31.099297046661377
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 1.155410871273134
P50 ttft = 0.8875744342803955
P99 ttft = 2.8498321056365974
Average tbt = 1.9389055775433055
P50 tbt = 1.7302134513854983
P99 tbt = 5.307018918991091
All GPUs and memories are cold after 36.09656858444214
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 2.468450665473938
P50 ttft = 2.5666087865829468
P99 ttft = 4.328714768886566
Average tbt = 7.615072207927705
P50 tbt = 7.602135765552522
P99 tbt = 15.121326914787295
All GPUs and memories are cold after 43.11873412132263
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 4.601233020424843
P50 ttft = 5.010060548782349
P99 ttft = 8.80265341758728
Average tbt = 9.770876002684236
P50 tbt = 9.733781468868258
P99 tbt = 19.3031876783371
All GPUs and memories are cold after 42.067432165145874
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 6.652650839661899
P50 ttft = 6.953061580657959
P99 ttft = 12.936849098205567
Average tbt = 11.168330891165015
P50 tbt = 11.123445534706118
P99 tbt = 21.995875587463384
All GPUs and memories are cold after 47.140833139419556
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 8.849945720419827
P50 ttft = 8.62835431098938
P99 ttft = 16.936321415901183
Average tbt = 12.728223101776768
P50 tbt = 12.68686995506287
P99 tbt = 25.092899027347567
All GPUs and memories are cold after 52.07440233230591
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.542660653591156
P50 ttft = 0.44442737102508545
P99 ttft = 1.032383348941803
Average tbt = 0.14562325874964396
P50 tbt = 0.018588829040527347
P99 tbt = 0.5969872217178348
All GPUs and memories are cold after 3.079155683517456
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.6515203544071743
P50 ttft = 0.5091879367828369
P99 ttft = 1.3537430763244631
Average tbt = 0.351975493204026
P50 tbt = 0.32517945766448986
P99 tbt = 1.1817053127288824
All GPUs and memories are cold after 12.024150371551514
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.951016582761492
P50 ttft = 0.7875368595123291
P99 ttft = 1.9710391712188708
Average tbt = 0.9356356757027764
P50 tbt = 0.6325273513793948
P99 tbt = 3.6125417275428773
All GPUs and memories are cold after 25.039082050323486
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 1.1359714647618735
P50 ttft = 0.8806416988372803
P99 ttft = 2.8074222087860115
Average tbt = 1.7721607283848089
P50 tbt = 1.6863421440124515
P99 tbt = 5.1137480211257955
All GPUs and memories are cold after 29.08054828643799
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 2.3919758558273316
P50 ttft = 2.4888012409210205
P99 ttft = 4.251736168861389
Average tbt = 7.60463047170639
P50 tbt = 7.600072598457339
P99 tbt = 15.054881666183476
All GPUs and memories are cold after 33.06721806526184
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 4.564235303550959
P50 ttft = 4.9712817668914795
P99 ttft = 8.76384259700775
Average tbt = 9.744225483015182
P50 tbt = 9.712273502349856
P99 tbt = 19.2527529706955
All GPUs and memories are cold after 38.062644243240356
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 6.71965712063933
P50 ttft = 7.032828092575073
P99 ttft = 13.058894243240356
Average tbt = 11.185301655612578
P50 tbt = 11.14091024398804
P99 tbt = 22.06684213542939
All GPUs and memories are cold after 45.06510543823242
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 8.84836770253009
P50 ttft = 8.621479749679565
P99 ttft = 16.946042203903197
Average tbt = 12.735468964691629
P50 tbt = 12.691501450538638
P99 tbt = 25.09924873399735
All GPUs and memories are cold after 44.07744860649109
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.542698880036672
P50 ttft = 0.4457831382751465
P99 ttft = 1.0300597667694094
Average tbt = 0.14541172186533616
P50 tbt = 0.018663716316223142
P99 tbt = 0.5950241737365726
All GPUs and memories are cold after 3.0124332904815674
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.6513626461937314
P50 ttft = 0.5067880153656006
P99 ttft = 1.3531611919403077
Average tbt = 0.35122302486783
P50 tbt = 0.32355427742004406
P99 tbt = 1.1806887149810796
All GPUs and memories are cold after 12.084543943405151
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.9433157307761056
P50 ttft = 0.7830076217651367
P99 ttft = 1.951262798309325
Average tbt = 0.9308495385306225
P50 tbt = 0.629124140739441
P99 tbt = 3.596123682975769
All GPUs and memories are cold after 25.040179014205933
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 1.1325094234652635
P50 ttft = 0.8813819885253906
P99 ttft = 2.804231977462769
Average tbt = 1.769767816473798
P50 tbt = 1.68005166053772
P99 tbt = 5.110402750968935
All GPUs and memories are cold after 34.07326126098633
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 2.2785286569595335
P50 ttft = 2.3760619163513184
P99 ttft = 4.125543882846832
Average tbt = 7.580269035339358
P50 tbt = 7.546618747711183
P99 tbt = 14.954123827934268
All GPUs and memories are cold after 35.05072474479675
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 4.561358794569969
P50 ttft = 4.870661497116089
P99 ttft = 8.836884608268736
Average tbt = 9.793153407052161
P50 tbt = 9.829882335662845
P99 tbt = 19.29941944432259
All GPUs and memories are cold after 40.06175088882446
92886
288 - huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000001:00:00.0
Warning: persistence mode is disabled on device 00000001:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000002:00:00.0
Warning: persistence mode is disabled on device 00000002:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000003:00:00.0
Warning: persistence mode is disabled on device 00000003:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000008:00:00.0
Warning: persistence mode is disabled on device 00000008:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 00000009:00:00.0
Warning: persistence mode is disabled on device 00000009:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000A:00:00.0
Warning: persistence mode is disabled on device 0000000A:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000B:00:00.0
Warning: persistence mode is disabled on device 0000000B:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
GPU clocks set to "(gpuClkMin 1980, gpuClkMax 1980)" for GPU 0000000C:00:00.0
Warning: persistence mode is disabled on device 0000000C:00:00.0. See the Known Issues section of the nvidia-smi(1) man page for more information. Run with [--help | -h] switch to get more information on how to enable persistence mode.
All done.
All done.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...
To disable this warning, you can either:
- Avoid using `tokenizers` before the fork if possible
- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 6.648875403077635
P50 ttft = 6.892650365829468
P99 ttft = 12.914585418701172
Average tbt = 11.13689418361612
P50 tbt = 11.136236381530765
P99 tbt = 21.961753652572636
All GPUs and memories are cold after 44.067726612091064
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 8.858141646327743
P50 ttft = 8.578660249710083
P99 ttft = 16.877743353843687
Average tbt = 12.6903696134866
P50 tbt = 12.679029965400698
P99 tbt = 25.05678614997864
All GPUs and memories are cold after 44.063825368881226
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.5599946777025858
P50 ttft = 0.44534242153167725
P99 ttft = 1.1344224095344544
Average tbt = 0.153014749288559
P50 tbt = 0.018639457225799558
P99 tbt = 0.6618595073223118
All GPUs and memories are cold after 4.038244009017944
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.6531296457563128
P50 ttft = 0.5093643665313721
P99 ttft = 1.357837724685669
Average tbt = 0.3519399733770462
P50 tbt = 0.3249937534332276
P99 tbt = 1.1826541137695317
All GPUs and memories are cold after 16.029732942581177
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.9472718375069754
P50 ttft = 0.7852139472961426
P99 ttft = 1.964170660972594
Average tbt = 0.9337384816578458
P50 tbt = 0.6311904668807985
P99 tbt = 3.607289986133575
All GPUs and memories are cold after 29.087682962417603
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 1.1402514271619844
P50 ttft = 0.883660078048706
P99 ttft = 2.8190935134887702
Average tbt = 1.775571512594456
P50 tbt = 1.6875597238540654
P99 tbt = 5.125080327987673
All GPUs and memories are cold after 34.07569241523743
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 2.3415367794036865
P50 ttft = 2.457382321357727
P99 ttft = 4.17460312128067
Average tbt = 7.578588411331177
P50 tbt = 7.56326484680176
P99 tbt = 14.999108510255818
All GPUs and memories are cold after 40.14042568206787
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 4.57400069758296
P50 ttft = 4.986700892448425
P99 ttft = 8.821406257152557
Average tbt = 9.777591733261943
P50 tbt = 9.736223661899569
P99 tbt = 19.290546898841864
All GPUs and memories are cold after 44.080082416534424
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 6.691866740788499
P50 ttft = 6.999942779541016
P99 ttft = 13.004151420593262
Average tbt = 11.168230865426265
P50 tbt = 11.12210068702698
P99 tbt = 22.03086738014222
All GPUs and memories are cold after 44.07079887390137
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 8.82846583803016
P50 ttft = 8.60572361946106
P99 ttft = 16.915447516441343
Average tbt = 12.73162910880813
P50 tbt = 12.683087658882144
P99 tbt = 25.079707905769354
All GPUs and memories are cold after 48.077595233917236
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.5442794760068258
P50 ttft = 0.4463711977005005
P99 ttft = 1.0324722075462343
Average tbt = 0.14584286411603295
P50 tbt = 0.018224906921386716
P99 tbt = 0.5964647727012637
All GPUs and memories are cold after 4.043173551559448
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600
8170 - 5
Average ttft = 0.6513547216142926
P50 ttft = 0.5093989372253418
P99 ttft = 1.3514277935028078
Average tbt = 0.35169806253342417
P50 tbt = 0.324325180053711
P99 tbt = 1.1816642475128178
All GPUs and memories are cold after 14.040027379989624
92886
288 - 96
288 - 256
288 - 1024
1053 - 96
1053 - 256
1053 - 600
8170 - 5
8170 - 256
8170 - 600