Hugging Face's logo

Diogo-V
/
opt-nlr-cosine-a18eef3dc781

English
Model card Files Community
Diogo-V commited on
Commit
6c7580c
·
verified ·
1 Parent(s): da4b605

Adds final graphs

Browse files
Files changed (43) hide show
  1. results/best_param_per_func_with_baselines/opt/4/best_param_per_func_with_baselines.png +0 -0
  2. results/best_param_per_func_with_baselines/opt/5/best_param_per_func_with_baselines.png +0 -0
  3. results/best_param_per_func_with_baselines/opt/6/best_param_per_func_with_baselines.png +0 -0
  4. results/best_param_per_func_with_baselines/opt/7/best_param_per_func_with_baselines.png +0 -0
  5. results/best_param_per_func_with_baselines/opt/8/best_param_per_func_with_baselines.png +0 -0
  6. results/bitwidth_acc_decay_per_function/opt.png +0 -0
  7. results/df.tsv +30 -77
  8. results/func_by_param_init_method/opt/4/comb_11.png +0 -0
  9. results/func_by_param_init_method/opt/4/comb_15.png +0 -0
  10. results/func_by_param_init_method/opt/4/comb_18.png +0 -0
  11. results/func_by_param_init_method/opt/4/comb_8.png +0 -0
  12. results/func_by_param_init_method/opt/5/comb_11.png +0 -0
  13. results/func_by_param_init_method/opt/5/comb_15.png +0 -0
  14. results/func_by_param_init_method/opt/5/comb_18.png +0 -0
  15. results/func_by_param_init_method/opt/5/comb_2.png +0 -0
  16. results/func_by_param_init_method/opt/5/comb_8.png +0 -0
  17. results/func_by_param_init_method/opt/6/comb_11.png +0 -0
  18. results/func_by_param_init_method/opt/6/comb_15.png +0 -0
  19. results/func_by_param_init_method/opt/6/comb_18.png +0 -0
  20. results/func_by_param_init_method/opt/6/comb_2.png +0 -0
  21. results/func_by_param_init_method/opt/6/comb_5.png +0 -0
  22. results/func_by_param_init_method/opt/6/comb_8.png +0 -0
  23. results/func_by_param_init_method/opt/7/comb_11.png +0 -0
  24. results/func_by_param_init_method/opt/7/comb_15.png +0 -0
  25. results/func_by_param_init_method/opt/7/comb_16.png +0 -0
  26. results/func_by_param_init_method/opt/7/comb_17.png +0 -0
  27. results/func_by_param_init_method/opt/7/comb_18.png +0 -0
  28. results/func_by_param_init_method/opt/7/comb_2.png +0 -0
  29. results/func_by_param_init_method/opt/7/comb_5.png +0 -0
  30. results/func_by_param_init_method/opt/7/comb_8.png +0 -0
  31. results/func_by_param_init_method/opt/7/comb_9.png +0 -0
  32. results/func_by_param_init_method/opt/8/comb_1.png +0 -0
  33. results/func_by_param_init_method/opt/8/comb_11.png +0 -0
  34. results/func_by_param_init_method/opt/8/comb_15.png +0 -0
  35. results/func_by_param_init_method/opt/8/comb_16.png +0 -0
  36. results/func_by_param_init_method/opt/8/comb_17.png +0 -0
  37. results/func_by_param_init_method/opt/8/comb_18.png +0 -0
  38. results/func_by_param_init_method/opt/8/comb_2.png +0 -0
  39. results/func_by_param_init_method/opt/8/comb_4.png +0 -0
  40. results/func_by_param_init_method/opt/8/comb_5.png +0 -0
  41. results/func_by_param_init_method/opt/8/comb_8.png +0 -0
  42. results/func_by_param_init_method/opt/8/comb_9.png +0 -0
  43. results/results.txt +71 -118
results/best_param_per_func_with_baselines/opt/4/best_param_per_func_with_baselines.png CHANGED
results/best_param_per_func_with_baselines/opt/5/best_param_per_func_with_baselines.png CHANGED
results/best_param_per_func_with_baselines/opt/6/best_param_per_func_with_baselines.png CHANGED
results/best_param_per_func_with_baselines/opt/7/best_param_per_func_with_baselines.png CHANGED
results/best_param_per_func_with_baselines/opt/8/best_param_per_func_with_baselines.png CHANGED
results/bitwidth_acc_decay_per_function/opt.png CHANGED
results/df.tsv CHANGED
@@ -5,13 +5,13 @@ accuracy bits clip dataset method model params peak_mem_alloc_mb peak_mem_reserv
5
  0.0 6 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
6
  0.0 7 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
7
  0.0 8 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
8
- 0.32445117831230164 8 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 840.957952 7.918829441070557 51.335494 0.0
9
- 0.32482269406318665 8 0.0 wikitext ao_woq opt fixed 580.071936 851.443712 7.916566848754883 34.694743 0.0
10
- 0.0031746034510433674 6 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.111747145652771 33.67688 0.0
11
  0.0 4 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.112660527229309 33.701441 0.0
12
- 0.0015873017255216837 7 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.1113218069076538 34.510958 0.0
13
- 0.0015873017255216837 5 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.113356113433838 35.914076 0.0
14
- 0.0015873017255216837 8 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.110521912574768 41.675911 0.0
15
  0.0 4 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
16
  0.0 5 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
17
  0.0 6 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
@@ -20,14 +20,14 @@ accuracy bits clip dataset method model params peak_mem_alloc_mb peak_mem_reserv
20
  0.318912535905838 5 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 8.089607238769531 27.840764 0.0
21
  0.3237757682800293 6 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.785684108734131 32.598827 0.0
22
  0.3243836760520935 8 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.910743713378906 33.10603 0.0
23
- 0.32519420981407166 7 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.901458740234375 33.579901 0.0
24
  0.3068895637989044 4 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 8.28001880645752 35.198885 0.0
25
- 0.31702131032943726 5 0.0 wikitext woq opt fixed 670.234624 861.929472 8.222683906555176 30.683926 0.0
26
  0.321073979139328 6 0.0 wikitext woq opt fixed 670.234624 861.929472 7.8056864738464355 31.461412 0.0
27
  0.3248564898967743 8 0.0 wikitext woq opt fixed 670.234624 861.929472 7.916111469268799 35.371139 0.0
28
  0.2924012243747711 4 0.0 wikitext woq opt fixed 670.234624 861.929472 8.903441429138184 36.692574 0.0
29
- 0.32451874017715454 7 0.0 wikitext woq opt fixed 670.234624 861.929472 7.905444145202637 38.010023 0.0
30
- 0.0015873017255216837 8 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1090807914733887 43.122991 0.0
31
  0.0 4 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
32
  0.0 5 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
33
  0.0 6 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
@@ -35,83 +35,36 @@ accuracy bits clip dataset method model params peak_mem_alloc_mb peak_mem_reserv
35
  0.0 5 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
36
  0.0 6 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
37
  0.0 7 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
38
- 0.01269841380417347 7 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.1033821105957031 47.727534 0.0
39
  0.0 4 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
40
  0.0 5 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
41
  0.0 6 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
42
- 0.32549813389778137 7 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 7.911805629730225 43.371287 0.0
43
  0.3072948157787323 4 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 8.315207481384277 43.593391 0.0
44
- 0.32127657532691956 6 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.803115367889404 43.770696 0.0
45
- 0.32489025592803955 7 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.927796840667725 45.935689 0.0
46
- 0.2889564335346222 8 0.0 wikitext comb_16 opt nlr_t_cos 671.452672 861.929472 9.226305961608887 47.901702 0.0
47
- 0.32529547810554504 7 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.930842876434326 48.020383 0.0
48
  0.3243836760520935 7 0.0 wikitext comb_17 opt nlr_t_cos 671.452672 861.929472 7.940777778625488 48.114679 0.0
49
  0.3184735178947449 5 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 8.065112113952637 48.629052 0.0
50
- 0.32441744208335876 8 0.0 wikitext comb_17 opt nlr_t_cos 671.452672 861.929472 7.916334629058838 48.773082 0.0
51
  0.3092873990535736 4 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 8.197556495666504 49.033408 0.0
52
  0.317325234413147 5 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 8.024106979370117 49.066547 0.0
53
- 0.32445117831230164 8 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.9176459312438965 49.377526 0.0
54
  0.325531929731369 7 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.898444652557373 49.692605 0.0
55
  0.3222559988498688 6 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 7.791579723358154 49.805036 0.0
56
- 0.32228976488113403 6 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.797433376312256 50.084441 0.0
57
- 0.31803444027900696 8 0.0 wikitext comb_9 opt nlr_t_cos 671.452672 861.929472 7.951894760131836 50.977965 0.0
58
  0.3186085820198059 5 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 8.072864532470703 52.201289 0.0
59
- 0.32158052921295166 6 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.811495304107666 52.832239 0.0
60
- 0.32404595613479614 8 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.927555084228516 54.346039 0.0
61
  0.1779804229736328 7 0.0 wikitext comb_9 opt nlr_t_cos 671.452672 861.929472 19.797229766845703 55.148583 0.0
62
- 0.32279637455940247 8 0.0 wikitext comb_1 opt nlr_t_cos 671.452672 861.929472 7.968763828277588 55.237912 0.0
63
- 0.32451874017715454 8 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.89610481262207 55.774946 0.0
64
- 0.30756503343582153 4 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 8.318281173706055 59.324934 0.0
65
- 0.0 8 0.0 wikitext comb_7_t opt nlr_t_cos 679.5136 861.929472 1.1143286228179932 51.694332 0.0
66
- 0.4351232945919037 8 0.0 wikitext comb_18_t opt nlr_t_cos 679.98464 861.929472 4.624080657958984 47.604596 0.0
67
- 0.0238095261156559 6 0.0 wikitext comb_2_t opt nlr_t_cos 678.220288 876.609536 1.0964152812957764 32.651327 0.0
68
- 0.01746032014489174 8 0.0 wikitext comb_2_t opt nlr_t_cos 678.220288 876.609536 1.097031593322754 33.026268 0.0
69
- 0.01904762163758278 4 0.0 wikitext comb_2_t opt nlr_t_cos 678.220288 876.609536 1.0909334421157837 33.323816 0.0
70
- 0.01428571529686451 5 0.0 wikitext comb_2_t opt nlr_t_cos 678.220288 876.609536 1.095602035522461 35.213413 0.0
71
- 0.02222222462296486 7 0.0 wikitext comb_2_t opt nlr_t_cos 678.220288 876.609536 1.097251296043396 41.142385 0.0
72
- 0.39341437816619873 6 0.0 wikitext comb_15_t opt nlr_t_cos 678.691328 876.609536 5.224710941314697 28.133357 0.0
73
- 0.43323200941085815 8 0.0 wikitext comb_15_t opt nlr_t_cos 678.691328 876.609536 4.6438446044921875 28.484168 0.0
74
- 0.36811888217926025 5 0.0 wikitext comb_15_t opt nlr_t_cos 678.691328 876.609536 5.857174873352051 31.667995 0.0
75
- 0.4268152713775635 7 0.0 wikitext comb_15_t opt nlr_t_cos 678.691328 876.609536 4.778627872467041 33.29018 0.0
76
- 0.3502195179462433 4 0.0 wikitext comb_15_t opt nlr_t_cos 678.691328 876.609536 6.451162338256836 33.816567 0.0
77
- 0.42992234230041504 8 0.0 wikitext woq_t opt fixed 678.766592 876.609536 4.648865222930908 30.294455 0.0
78
- 0.4171901345252991 7 0.0 wikitext woq_t opt fixed 678.766592 876.609536 4.781289100646973 30.536944 0.0
79
- 0.3299223482608795 4 0.0 wikitext woq_t opt fixed 678.766592 876.609536 6.8836565017700195 31.971951 0.0
80
- 0.36555215716362 5 0.0 wikitext woq_t opt fixed 678.766592 876.609536 6.027744770050049 33.029717 0.0
81
- 0.3956433832645416 6 0.0 wikitext woq_t opt fixed 678.766592 876.609536 5.197286128997803 33.071645 0.0
82
- 0.0 8 0.0 wikitext comb_4_t opt nlr_t_cos 679.5136 876.609536 1.0890827178955078 47.159053 0.0
83
- 0.0031746034510433674 7 0.0 wikitext comb_4_t opt nlr_t_cos 679.5136 876.609536 1.1006628274917603 52.275743 0.0
84
- 0.03439153730869293 6 0.0 wikitext comb_9_t opt nlr_t_cos 679.515648 876.609536 1.0895891189575195 49.916958 0.0
85
- 0.03439153730869293 8 0.0 wikitext comb_9_t opt nlr_t_cos 679.515648 876.609536 1.0900154113769531 53.497639 0.0
86
- 0.03439153730869293 7 0.0 wikitext comb_9_t opt nlr_t_cos 679.515648 876.609536 1.0896639823913574 65.132813 0.0
87
- 0.43515706062316895 8 0.0 wikitext comb_11_t opt nlr_t_cos 679.98464 876.609536 4.593360424041748 43.405059 0.0
88
- 0.3704153895378113 5 0.0 wikitext comb_11_t opt nlr_t_cos 679.98464 876.609536 5.741970062255859 45.853767 0.0
89
- 0.35086119174957275 4 0.0 wikitext comb_11_t opt nlr_t_cos 679.98464 876.609536 6.511904716491699 48.820796 0.0
90
- 0.3951367735862732 6 0.0 wikitext comb_11_t opt nlr_t_cos 679.98464 876.609536 5.198108196258545 49.086005 0.0
91
- 0.42174941301345825 8 0.0 wikitext comb_8_t opt nlr_t_cos 679.98464 876.609536 4.704576015472412 50.212442 0.0
92
- 0.4170888662338257 7 0.0 wikitext comb_8_t opt nlr_t_cos 679.98464 876.609536 4.795653820037842 50.727664 0.0
93
- 0.40486326813697815 6 0.0 wikitext comb_8_t opt nlr_t_cos 679.98464 876.609536 5.209366321563721 54.19264 0.0
94
- 0.41519758105278015 7 0.0 wikitext comb_11_t opt nlr_t_cos 679.98464 876.609536 4.9116597175598145 56.004354 0.0
95
- 0.35069236159324646 4 0.0 wikitext comb_8_t opt nlr_t_cos 679.98464 876.609536 6.456235408782959 57.128501 0.0
96
- 0.3715974688529968 5 0.0 wikitext comb_8_t opt nlr_t_cos 679.98464 876.609536 5.772923946380615 57.204256 0.0
97
- 0.0 8 0.0 wikitext comb_6_t opt nlr_t_cos 678.220288 882.900992 1.1143286228179932 39.20088 0.0
98
- 0.03439153730869293 8 0.0 wikitext comb_16_t opt nlr_t_cos 679.515648 882.900992 1.0895816087722778 42.991725 0.0
99
- 0.03439153730869293 7 0.0 wikitext comb_16_t opt nlr_t_cos 679.515648 882.900992 1.090133547782898 43.597738 0.0
100
- 0.03439153730869293 6 0.0 wikitext comb_16_t opt nlr_t_cos 679.515648 882.900992 1.0899640321731567 48.70578 0.0
101
- 0.4039851427078247 6 0.0 wikitext comb_18_t opt nlr_t_cos 679.98464 882.900992 5.196511268615723 44.049948 0.0
102
- 0.4176967442035675 7 0.0 wikitext comb_18_t opt nlr_t_cos 679.98464 882.900992 4.876777172088623 45.574922 0.0
103
- 0.3756163418292999 5 0.0 wikitext comb_18_t opt nlr_t_cos 679.98464 882.900992 5.780261516571045 45.886395 0.0
104
- 0.4369807541370392 8 0.0 wikitext comb_1_t opt nlr_t_cos 679.98464 882.900992 4.652184963226318 45.947897 0.0
105
- 0.4223910868167877 8 0.0 wikitext comb_5_t opt nlr_t_cos 679.98464 882.900992 4.792054653167725 47.138481 0.0
106
- 0.3971293568611145 6 0.0 wikitext comb_5_t opt nlr_t_cos 679.98464 882.900992 5.220573425292969 47.303211 0.0
107
- 0.3511313796043396 4 0.0 wikitext comb_18_t opt nlr_t_cos 679.98464 882.900992 6.439454555511475 51.203371 0.0
108
- 0.42522796988487244 8 0.0 wikitext comb_17_t opt nlr_t_cos 679.98464 882.900992 4.6705427169799805 52.309068 0.0
109
- 0.41789939999580383 7 0.0 wikitext comb_5_t opt nlr_t_cos 679.98464 882.900992 4.7880964279174805 55.913264 0.0
110
- 0.42198583483695984 7 0.0 wikitext comb_17_t opt nlr_t_cos 679.98464 882.900992 4.843860626220703 57.694705 0.0
111
- 0.0 4 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.1121183633804321 0.071884 0.0
112
- 0.0 5 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.1121183633804321 0.071884 0.0
113
- 0.0 6 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.1121183633804321 0.071884 0.0
114
- 0.0 7 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.1121183633804321 0.071884 0.0
115
- 0.0 8 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.1121183633804321 0.071884 0.0
116
- 0.0 8 0.0 wikitext comb_14_t opt nlr_t_cos 486.259712 1073.741824 1.1089136600494385 0.09668 0.0
117
  0.3243836760520935 32 0.0 wikitext base opt ---- 1359.398912 1402.994688 7.934948921203613 19.520466 0.0
 
5
  0.0 6 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
6
  0.0 7 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
7
  0.0 8 0.0 wikitext comb_7 opt nlr_t_cos 670.981632 840.957952 1.1143286228179932 55.137657 0.0
8
+ 0.3244511783123016 8 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 840.957952 7.918829441070557 51.335494 0.0
9
+ 0.3248226940631866 8 0.0 wikitext ao_woq opt fixed 580.071936 851.443712 7.916566848754883 34.694743 0.0
10
+ 0.0031746034510433 6 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.111747145652771 33.67688 0.0
11
  0.0 4 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.112660527229309 33.701441 0.0
12
+ 0.0015873017255216 7 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.1113218069076538 34.510958 0.0
13
+ 0.0015873017255216 5 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.113356113433838 35.914076 0.0
14
+ 0.0015873017255216 8 0.0 wikitext comb_2 opt nlr_t_cos 669.68832 861.929472 1.110521912574768 41.675911 0.0
15
  0.0 4 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
16
  0.0 5 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
17
  0.0 6 0.0 wikitext comb_6 opt nlr_t_cos 669.68832 861.929472 1.1143286228179932 43.23713 0.0
 
20
  0.318912535905838 5 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 8.089607238769531 27.840764 0.0
21
  0.3237757682800293 6 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.785684108734131 32.598827 0.0
22
  0.3243836760520935 8 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.910743713378906 33.10603 0.0
23
+ 0.3251942098140716 7 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 7.901458740234375 33.579901 0.0
24
  0.3068895637989044 4 0.0 wikitext comb_15 opt nlr_t_cos 670.15936 861.929472 8.28001880645752 35.198885 0.0
25
+ 0.3170213103294372 5 0.0 wikitext woq opt fixed 670.234624 861.929472 8.222683906555176 30.683926 0.0
26
  0.321073979139328 6 0.0 wikitext woq opt fixed 670.234624 861.929472 7.8056864738464355 31.461412 0.0
27
  0.3248564898967743 8 0.0 wikitext woq opt fixed 670.234624 861.929472 7.916111469268799 35.371139 0.0
28
  0.2924012243747711 4 0.0 wikitext woq opt fixed 670.234624 861.929472 8.903441429138184 36.692574 0.0
29
+ 0.3245187401771545 7 0.0 wikitext woq opt fixed 670.234624 861.929472 7.905444145202637 38.010023 0.0
30
+ 0.0015873017255216 8 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.109080791473389 43.122991 0.0
31
  0.0 4 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
32
  0.0 5 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
33
  0.0 6 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.115140676498413 45.188353 0.0
 
35
  0.0 5 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
36
  0.0 6 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
37
  0.0 7 0.0 wikitext comb_4 opt nlr_t_cos 670.981632 861.929472 1.1103847026824951 45.260816 0.0
38
+ 0.0126984138041734 7 0.0 wikitext comb_16 opt nlr_t_cos 670.981632 861.929472 1.1033821105957031 47.727534 0.0
39
  0.0 4 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
40
  0.0 5 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
41
  0.0 6 0.0 wikitext comb_9 opt nlr_t_cos 670.981632 861.929472 1.115138053894043 53.780705 0.0
42
+ 0.3254981338977813 7 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 7.911805629730225 43.371287 0.0
43
  0.3072948157787323 4 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 8.315207481384277 43.593391 0.0
44
+ 0.3212765753269195 6 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.803115367889404 43.770696 0.0
45
+ 0.3248902559280395 7 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.927796840667725 45.935689 0.0
46
+ 0.2889564335346222 8 0.0 wikitext comb_16 opt nlr_t_cos 671.452672 861.929472 9.226305961608888 47.901702 0.0
47
+ 0.325295478105545 7 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.930842876434326 48.020383 0.0
48
  0.3243836760520935 7 0.0 wikitext comb_17 opt nlr_t_cos 671.452672 861.929472 7.940777778625488 48.114679 0.0
49
  0.3184735178947449 5 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 8.065112113952637 48.629052 0.0
50
+ 0.3244174420833587 8 0.0 wikitext comb_17 opt nlr_t_cos 671.452672 861.929472 7.916334629058838 48.773082 0.0
51
  0.3092873990535736 4 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 8.197556495666504 49.033408 0.0
52
  0.317325234413147 5 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 8.024106979370117 49.066547 0.0
53
+ 0.3244511783123016 8 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.917645931243896 49.377526 0.0
54
  0.325531929731369 7 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.898444652557373 49.692605 0.0
55
  0.3222559988498688 6 0.0 wikitext comb_18 opt nlr_t_cos 671.452672 861.929472 7.791579723358154 49.805036 0.0
56
+ 0.322289764881134 6 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 7.797433376312256 50.084441 0.0
57
+ 0.3180344402790069 8 0.0 wikitext comb_9 opt nlr_t_cos 671.452672 861.929472 7.951894760131836 50.977965 0.0
58
  0.3186085820198059 5 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 8.072864532470703 52.201289 0.0
59
+ 0.3215805292129516 6 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.811495304107666 52.832239 0.0
60
+ 0.3240459561347961 8 0.0 wikitext comb_5 opt nlr_t_cos 671.452672 861.929472 7.927555084228516 54.346039 0.0
61
  0.1779804229736328 7 0.0 wikitext comb_9 opt nlr_t_cos 671.452672 861.929472 19.797229766845703 55.148583 0.0
62
+ 0.3227963745594024 8 0.0 wikitext comb_1 opt nlr_t_cos 671.452672 861.929472 7.968763828277588 55.237912 0.0
63
+ 0.3245187401771545 8 0.0 wikitext comb_11 opt nlr_t_cos 671.452672 861.929472 7.89610481262207 55.774946 0.0
64
+ 0.3075650334358215 4 0.0 wikitext comb_8 opt nlr_t_cos 671.452672 861.929472 8.318281173706055 59.324934 0.0
65
+ 0.0 4 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.112118363380432 0.071884 0.0
66
+ 0.0 5 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.112118363380432 0.071884 0.0
67
+ 0.0 6 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.112118363380432 0.071884 0.0
68
+ 0.0 7 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.112118363380432 0.071884 0.0
69
+ 0.0 8 0.0 wikitext comb_14 opt nlr_t_cos 477.740032 968.884224 1.112118363380432 0.071884 0.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
70
  0.3243836760520935 32 0.0 wikitext base opt ---- 1359.398912 1402.994688 7.934948921203613 19.520466 0.0
results/func_by_param_init_method/opt/4/comb_11.png CHANGED
results/func_by_param_init_method/opt/4/comb_15.png CHANGED
results/func_by_param_init_method/opt/4/comb_18.png CHANGED
results/func_by_param_init_method/opt/4/comb_8.png CHANGED
results/func_by_param_init_method/opt/5/comb_11.png CHANGED
results/func_by_param_init_method/opt/5/comb_15.png CHANGED
results/func_by_param_init_method/opt/5/comb_18.png CHANGED
results/func_by_param_init_method/opt/5/comb_2.png CHANGED
results/func_by_param_init_method/opt/5/comb_8.png CHANGED
results/func_by_param_init_method/opt/6/comb_11.png CHANGED
results/func_by_param_init_method/opt/6/comb_15.png CHANGED
results/func_by_param_init_method/opt/6/comb_18.png CHANGED
results/func_by_param_init_method/opt/6/comb_2.png CHANGED
results/func_by_param_init_method/opt/6/comb_5.png CHANGED
results/func_by_param_init_method/opt/6/comb_8.png CHANGED
results/func_by_param_init_method/opt/7/comb_11.png CHANGED
results/func_by_param_init_method/opt/7/comb_15.png CHANGED
results/func_by_param_init_method/opt/7/comb_16.png CHANGED
results/func_by_param_init_method/opt/7/comb_17.png CHANGED
results/func_by_param_init_method/opt/7/comb_18.png CHANGED
results/func_by_param_init_method/opt/7/comb_2.png CHANGED
results/func_by_param_init_method/opt/7/comb_5.png CHANGED
results/func_by_param_init_method/opt/7/comb_8.png CHANGED
results/func_by_param_init_method/opt/7/comb_9.png CHANGED
results/func_by_param_init_method/opt/8/comb_1.png CHANGED
results/func_by_param_init_method/opt/8/comb_11.png CHANGED
results/func_by_param_init_method/opt/8/comb_15.png CHANGED
results/func_by_param_init_method/opt/8/comb_16.png CHANGED
results/func_by_param_init_method/opt/8/comb_17.png CHANGED
results/func_by_param_init_method/opt/8/comb_18.png CHANGED
results/func_by_param_init_method/opt/8/comb_2.png CHANGED
results/func_by_param_init_method/opt/8/comb_4.png CHANGED
results/func_by_param_init_method/opt/8/comb_5.png CHANGED
results/func_by_param_init_method/opt/8/comb_8.png CHANGED
results/func_by_param_init_method/opt/8/comb_9.png CHANGED
results/results.txt CHANGED
@@ -1,118 +1,71 @@
1
- | | accuracy | bits | clip | dataset | method | model | params | peak_mem_alloc_mb | peak_mem_reserv_mb | perplexity | time_ms | wer |
2
- |----:|-----------:|-------:|-------:|:----------|:----------|:--------|:----------|--------------------:|---------------------:|-------------:|----------:|------:|
3
- | 0 | 0.296589 | 4 | 0 | wikitext | ao_woq | opt | fixed | 268.126 | 677.38 | 8.76863 | 37.2247 | 0 |
4
- | 1 | 0 | 4 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.11433 | 55.1377 | 0 |
5
- | 2 | 0 | 5 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.11433 | 55.1377 | 0 |
6
- | 3 | 0 | 6 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.11433 | 55.1377 | 0 |
7
- | 4 | 0 | 7 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.11433 | 55.1377 | 0 |
8
- | 5 | 0 | 8 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.11433 | 55.1377 | 0 |
9
- | 6 | 0.324451 | 8 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 840.958 | 7.91883 | 51.3355 | 0 |
10
- | 7 | 0.324823 | 8 | 0 | wikitext | ao_woq | opt | fixed | 580.072 | 851.444 | 7.91657 | 34.6947 | 0 |
11
- | 8 | 0.0031746 | 6 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11175 | 33.6769 | 0 |
12
- | 9 | 0 | 4 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11266 | 33.7014 | 0 |
13
- | 10 | 0.0015873 | 7 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11132 | 34.511 | 0 |
14
- | 11 | 0.0015873 | 5 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11336 | 35.9141 | 0 |
15
- | 12 | 0.0015873 | 8 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11052 | 41.6759 | 0 |
16
- | 13 | 0 | 4 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11433 | 43.2371 | 0 |
17
- | 14 | 0 | 5 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11433 | 43.2371 | 0 |
18
- | 15 | 0 | 6 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11433 | 43.2371 | 0 |
19
- | 16 | 0 | 7 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11433 | 43.2371 | 0 |
20
- | 17 | 0 | 8 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.11433 | 43.2371 | 0 |
21
- | 18 | 0.318913 | 5 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 8.08961 | 27.8408 | 0 |
22
- | 19 | 0.323776 | 6 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.78568 | 32.5988 | 0 |
23
- | 20 | 0.324384 | 8 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.91074 | 33.106 | 0 |
24
- | 21 | 0.325194 | 7 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.90146 | 33.5799 | 0 |
25
- | 22 | 0.30689 | 4 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 8.28002 | 35.1989 | 0 |
26
- | 23 | 0.317021 | 5 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 8.22268 | 30.6839 | 0 |
27
- | 24 | 0.321074 | 6 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.80569 | 31.4614 | 0 |
28
- | 25 | 0.324856 | 8 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.91611 | 35.3711 | 0 |
29
- | 26 | 0.292401 | 4 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 8.90344 | 36.6926 | 0 |
30
- | 27 | 0.324519 | 7 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.90544 | 38.01 | 0 |
31
- | 28 | 0.0015873 | 8 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.10908 | 43.123 | 0 |
32
- | 29 | 0 | 4 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 45.1884 | 0 |
33
- | 30 | 0 | 5 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 45.1884 | 0 |
34
- | 31 | 0 | 6 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 45.1884 | 0 |
35
- | 32 | 0 | 4 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11038 | 45.2608 | 0 |
36
- | 33 | 0 | 5 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11038 | 45.2608 | 0 |
37
- | 34 | 0 | 6 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11038 | 45.2608 | 0 |
38
- | 35 | 0 | 7 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11038 | 45.2608 | 0 |
39
- | 36 | 0.0126984 | 7 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.10338 | 47.7275 | 0 |
40
- | 37 | 0 | 4 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 53.7807 | 0 |
41
- | 38 | 0 | 5 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 53.7807 | 0 |
42
- | 39 | 0 | 6 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.11514 | 53.7807 | 0 |
43
- | 40 | 0.325498 | 7 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 7.91181 | 43.3713 | 0 |
44
- | 41 | 0.307295 | 4 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 8.31521 | 43.5934 | 0 |
45
- | 42 | 0.321277 | 6 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.80312 | 43.7707 | 0 |
46
- | 43 | 0.32489 | 7 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9278 | 45.9357 | 0 |
47
- | 44 | 0.288956 | 8 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 671.453 | 861.929 | 9.22631 | 47.9017 | 0 |
48
- | 45 | 0.325295 | 7 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.93084 | 48.0204 | 0 |
49
- | 46 | 0.324384 | 7 | 0 | wikitext | comb_17 | opt | nlr_t_cos | 671.453 | 861.929 | 7.94078 | 48.1147 | 0 |
50
- | 47 | 0.318474 | 5 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 8.06511 | 48.6291 | 0 |
51
- | 48 | 0.324417 | 8 | 0 | wikitext | comb_17 | opt | nlr_t_cos | 671.453 | 861.929 | 7.91633 | 48.7731 | 0 |
52
- | 49 | 0.309287 | 4 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 8.19756 | 49.0334 | 0 |
53
- | 50 | 0.317325 | 5 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 8.02411 | 49.0665 | 0 |
54
- | 51 | 0.324451 | 8 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.91765 | 49.3775 | 0 |
55
- | 52 | 0.325532 | 7 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.89844 | 49.6926 | 0 |
56
- | 53 | 0.322256 | 6 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 7.79158 | 49.805 | 0 |
57
- | 54 | 0.32229 | 6 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.79743 | 50.0844 | 0 |
58
- | 55 | 0.318034 | 8 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 671.453 | 861.929 | 7.95189 | 50.978 | 0 |
59
- | 56 | 0.318609 | 5 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 8.07286 | 52.2013 | 0 |
60
- | 57 | 0.321581 | 6 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8115 | 52.8322 | 0 |
61
- | 58 | 0.324046 | 8 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.92756 | 54.346 | 0 |
62
- | 59 | 0.17798 | 7 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 671.453 | 861.929 | 19.7972 | 55.1486 | 0 |
63
- | 60 | 0.322796 | 8 | 0 | wikitext | comb_1 | opt | nlr_t_cos | 671.453 | 861.929 | 7.96876 | 55.2379 | 0 |
64
- | 61 | 0.324519 | 8 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8961 | 55.7749 | 0 |
65
- | 62 | 0.307565 | 4 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 8.31828 | 59.3249 | 0 |
66
- | 63 | 0 | 8 | 0 | wikitext | comb_7_t | opt | nlr_t_cos | 679.514 | 861.929 | 1.11433 | 51.6943 | 0 |
67
- | 64 | 0.435123 | 8 | 0 | wikitext | comb_18_t | opt | nlr_t_cos | 679.985 | 861.929 | 4.62408 | 47.6046 | 0 |
68
- | 65 | 0.0238095 | 6 | 0 | wikitext | comb_2_t | opt | nlr_t_cos | 678.22 | 876.61 | 1.09642 | 32.6513 | 0 |
69
- | 66 | 0.0174603 | 8 | 0 | wikitext | comb_2_t | opt | nlr_t_cos | 678.22 | 876.61 | 1.09703 | 33.0263 | 0 |
70
- | 67 | 0.0190476 | 4 | 0 | wikitext | comb_2_t | opt | nlr_t_cos | 678.22 | 876.61 | 1.09093 | 33.3238 | 0 |
71
- | 68 | 0.0142857 | 5 | 0 | wikitext | comb_2_t | opt | nlr_t_cos | 678.22 | 876.61 | 1.0956 | 35.2134 | 0 |
72
- | 69 | 0.0222222 | 7 | 0 | wikitext | comb_2_t | opt | nlr_t_cos | 678.22 | 876.61 | 1.09725 | 41.1424 | 0 |
73
- | 70 | 0.393414 | 6 | 0 | wikitext | comb_15_t | opt | nlr_t_cos | 678.691 | 876.61 | 5.22471 | 28.1334 | 0 |
74
- | 71 | 0.433232 | 8 | 0 | wikitext | comb_15_t | opt | nlr_t_cos | 678.691 | 876.61 | 4.64384 | 28.4842 | 0 |
75
- | 72 | 0.368119 | 5 | 0 | wikitext | comb_15_t | opt | nlr_t_cos | 678.691 | 876.61 | 5.85717 | 31.668 | 0 |
76
- | 73 | 0.426815 | 7 | 0 | wikitext | comb_15_t | opt | nlr_t_cos | 678.691 | 876.61 | 4.77863 | 33.2902 | 0 |
77
- | 74 | 0.35022 | 4 | 0 | wikitext | comb_15_t | opt | nlr_t_cos | 678.691 | 876.61 | 6.45116 | 33.8166 | 0 |
78
- | 75 | 0.429922 | 8 | 0 | wikitext | woq_t | opt | fixed | 678.767 | 876.61 | 4.64887 | 30.2945 | 0 |
79
- | 76 | 0.41719 | 7 | 0 | wikitext | woq_t | opt | fixed | 678.767 | 876.61 | 4.78129 | 30.5369 | 0 |
80
- | 77 | 0.329922 | 4 | 0 | wikitext | woq_t | opt | fixed | 678.767 | 876.61 | 6.88366 | 31.972 | 0 |
81
- | 78 | 0.365552 | 5 | 0 | wikitext | woq_t | opt | fixed | 678.767 | 876.61 | 6.02774 | 33.0297 | 0 |
82
- | 79 | 0.395643 | 6 | 0 | wikitext | woq_t | opt | fixed | 678.767 | 876.61 | 5.19729 | 33.0716 | 0 |
83
- | 80 | 0 | 8 | 0 | wikitext | comb_4_t | opt | nlr_t_cos | 679.514 | 876.61 | 1.08908 | 47.1591 | 0 |
84
- | 81 | 0.0031746 | 7 | 0 | wikitext | comb_4_t | opt | nlr_t_cos | 679.514 | 876.61 | 1.10066 | 52.2757 | 0 |
85
- | 82 | 0.0343915 | 6 | 0 | wikitext | comb_9_t | opt | nlr_t_cos | 679.516 | 876.61 | 1.08959 | 49.917 | 0 |
86
- | 83 | 0.0343915 | 8 | 0 | wikitext | comb_9_t | opt | nlr_t_cos | 679.516 | 876.61 | 1.09002 | 53.4976 | 0 |
87
- | 84 | 0.0343915 | 7 | 0 | wikitext | comb_9_t | opt | nlr_t_cos | 679.516 | 876.61 | 1.08966 | 65.1328 | 0 |
88
- | 85 | 0.435157 | 8 | 0 | wikitext | comb_11_t | opt | nlr_t_cos | 679.985 | 876.61 | 4.59336 | 43.4051 | 0 |
89
- | 86 | 0.370415 | 5 | 0 | wikitext | comb_11_t | opt | nlr_t_cos | 679.985 | 876.61 | 5.74197 | 45.8538 | 0 |
90
- | 87 | 0.350861 | 4 | 0 | wikitext | comb_11_t | opt | nlr_t_cos | 679.985 | 876.61 | 6.5119 | 48.8208 | 0 |
91
- | 88 | 0.395137 | 6 | 0 | wikitext | comb_11_t | opt | nlr_t_cos | 679.985 | 876.61 | 5.19811 | 49.086 | 0 |
92
- | 89 | 0.421749 | 8 | 0 | wikitext | comb_8_t | opt | nlr_t_cos | 679.985 | 876.61 | 4.70458 | 50.2124 | 0 |
93
- | 90 | 0.417089 | 7 | 0 | wikitext | comb_8_t | opt | nlr_t_cos | 679.985 | 876.61 | 4.79565 | 50.7277 | 0 |
94
- | 91 | 0.404863 | 6 | 0 | wikitext | comb_8_t | opt | nlr_t_cos | 679.985 | 876.61 | 5.20937 | 54.1926 | 0 |
95
- | 92 | 0.415198 | 7 | 0 | wikitext | comb_11_t | opt | nlr_t_cos | 679.985 | 876.61 | 4.91166 | 56.0044 | 0 |
96
- | 93 | 0.350692 | 4 | 0 | wikitext | comb_8_t | opt | nlr_t_cos | 679.985 | 876.61 | 6.45624 | 57.1285 | 0 |
97
- | 94 | 0.371597 | 5 | 0 | wikitext | comb_8_t | opt | nlr_t_cos | 679.985 | 876.61 | 5.77292 | 57.2043 | 0 |
98
- | 95 | 0 | 8 | 0 | wikitext | comb_6_t | opt | nlr_t_cos | 678.22 | 882.901 | 1.11433 | 39.2009 | 0 |
99
- | 96 | 0.0343915 | 8 | 0 | wikitext | comb_16_t | opt | nlr_t_cos | 679.516 | 882.901 | 1.08958 | 42.9917 | 0 |
100
- | 97 | 0.0343915 | 7 | 0 | wikitext | comb_16_t | opt | nlr_t_cos | 679.516 | 882.901 | 1.09013 | 43.5977 | 0 |
101
- | 98 | 0.0343915 | 6 | 0 | wikitext | comb_16_t | opt | nlr_t_cos | 679.516 | 882.901 | 1.08996 | 48.7058 | 0 |
102
- | 99 | 0.403985 | 6 | 0 | wikitext | comb_18_t | opt | nlr_t_cos | 679.985 | 882.901 | 5.19651 | 44.0499 | 0 |
103
- | 100 | 0.417697 | 7 | 0 | wikitext | comb_18_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.87678 | 45.5749 | 0 |
104
- | 101 | 0.375616 | 5 | 0 | wikitext | comb_18_t | opt | nlr_t_cos | 679.985 | 882.901 | 5.78026 | 45.8864 | 0 |
105
- | 102 | 0.436981 | 8 | 0 | wikitext | comb_1_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.65218 | 45.9479 | 0 |
106
- | 103 | 0.422391 | 8 | 0 | wikitext | comb_5_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.79205 | 47.1385 | 0 |
107
- | 104 | 0.397129 | 6 | 0 | wikitext | comb_5_t | opt | nlr_t_cos | 679.985 | 882.901 | 5.22057 | 47.3032 | 0 |
108
- | 105 | 0.351131 | 4 | 0 | wikitext | comb_18_t | opt | nlr_t_cos | 679.985 | 882.901 | 6.43945 | 51.2034 | 0 |
109
- | 106 | 0.425228 | 8 | 0 | wikitext | comb_17_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.67054 | 52.3091 | 0 |
110
- | 107 | 0.417899 | 7 | 0 | wikitext | comb_5_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.7881 | 55.9133 | 0 |
111
- | 108 | 0.421986 | 7 | 0 | wikitext | comb_17_t | opt | nlr_t_cos | 679.985 | 882.901 | 4.84386 | 57.6947 | 0 |
112
- | 109 | 0 | 4 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.11212 | 0.071884 | 0 |
113
- | 110 | 0 | 5 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.11212 | 0.071884 | 0 |
114
- | 111 | 0 | 6 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.11212 | 0.071884 | 0 |
115
- | 112 | 0 | 7 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.11212 | 0.071884 | 0 |
116
- | 113 | 0 | 8 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.11212 | 0.071884 | 0 |
117
- | 114 | 0 | 8 | 0 | wikitext | comb_14_t | opt | nlr_t_cos | 486.26 | 1073.74 | 1.10891 | 0.09668 | 0 |
118
- | 115 | 0.324384 | 32 | 0 | wikitext | base | opt | ---- | 1359.4 | 1402.99 | 7.93495 | 19.5205 | 0 |
 
1
+ | | accuracy | bits | clip | dataset | method | model | params | peak_mem_alloc_mb | peak_mem_reserv_mb | perplexity | time_ms | wer |
2
+ |---:|-----------:|-------:|-------:|:----------|:---------|:--------|:----------|--------------------:|---------------------:|-------------:|----------:|------:|
3
+ | 0 | 0.2966 | 4 | 0 | wikitext | ao_woq | opt | fixed | 268.126 | 677.38 | 8.7686 | 37.2247 | 0 |
4
+ | 1 | 0 | 4 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.1143 | 55.1377 | 0 |
5
+ | 2 | 0 | 5 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.1143 | 55.1377 | 0 |
6
+ | 3 | 0 | 6 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.1143 | 55.1377 | 0 |
7
+ | 4 | 0 | 7 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.1143 | 55.1377 | 0 |
8
+ | 5 | 0 | 8 | 0 | wikitext | comb_7 | opt | nlr_t_cos | 670.982 | 840.958 | 1.1143 | 55.1377 | 0 |
9
+ | 6 | 0.3245 | 8 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 840.958 | 7.9188 | 51.3355 | 0 |
10
+ | 7 | 0.3248 | 8 | 0 | wikitext | ao_woq | opt | fixed | 580.072 | 851.444 | 7.9166 | 34.6947 | 0 |
11
+ | 8 | 0.0032 | 6 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1117 | 33.6769 | 0 |
12
+ | 9 | 0 | 4 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1127 | 33.7014 | 0 |
13
+ | 10 | 0.0016 | 7 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1113 | 34.511 | 0 |
14
+ | 11 | 0.0016 | 5 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1134 | 35.9141 | 0 |
15
+ | 12 | 0.0016 | 8 | 0 | wikitext | comb_2 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1105 | 41.6759 | 0 |
16
+ | 13 | 0 | 4 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1143 | 43.2371 | 0 |
17
+ | 14 | 0 | 5 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1143 | 43.2371 | 0 |
18
+ | 15 | 0 | 6 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1143 | 43.2371 | 0 |
19
+ | 16 | 0 | 7 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1143 | 43.2371 | 0 |
20
+ | 17 | 0 | 8 | 0 | wikitext | comb_6 | opt | nlr_t_cos | 669.688 | 861.929 | 1.1143 | 43.2371 | 0 |
21
+ | 18 | 0.3189 | 5 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 8.0896 | 27.8408 | 0 |
22
+ | 19 | 0.3238 | 6 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.7857 | 32.5988 | 0 |
23
+ | 20 | 0.3244 | 8 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.9107 | 33.106 | 0 |
24
+ | 21 | 0.3252 | 7 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 7.9015 | 33.5799 | 0 |
25
+ | 22 | 0.3069 | 4 | 0 | wikitext | comb_15 | opt | nlr_t_cos | 670.159 | 861.929 | 8.28 | 35.1989 | 0 |
26
+ | 23 | 0.317 | 5 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 8.2227 | 30.6839 | 0 |
27
+ | 24 | 0.3211 | 6 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.8057 | 31.4614 | 0 |
28
+ | 25 | 0.3249 | 8 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.9161 | 35.3711 | 0 |
29
+ | 26 | 0.2924 | 4 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 8.9034 | 36.6926 | 0 |
30
+ | 27 | 0.3245 | 7 | 0 | wikitext | woq | opt | fixed | 670.235 | 861.929 | 7.9054 | 38.01 | 0 |
31
+ | 28 | 0.0016 | 8 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1091 | 43.123 | 0 |
32
+ | 29 | 0 | 4 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 45.1884 | 0 |
33
+ | 30 | 0 | 5 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 45.1884 | 0 |
34
+ | 31 | 0 | 6 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 45.1884 | 0 |
35
+ | 32 | 0 | 4 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1104 | 45.2608 | 0 |
36
+ | 33 | 0 | 5 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1104 | 45.2608 | 0 |
37
+ | 34 | 0 | 6 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1104 | 45.2608 | 0 |
38
+ | 35 | 0 | 7 | 0 | wikitext | comb_4 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1104 | 45.2608 | 0 |
39
+ | 36 | 0.0127 | 7 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1034 | 47.7275 | 0 |
40
+ | 37 | 0 | 4 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 53.7807 | 0 |
41
+ | 38 | 0 | 5 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 53.7807 | 0 |
42
+ | 39 | 0 | 6 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 670.982 | 861.929 | 1.1151 | 53.7807 | 0 |
43
+ | 40 | 0.3255 | 7 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9118 | 43.3713 | 0 |
44
+ | 41 | 0.3073 | 4 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 8.3152 | 43.5934 | 0 |
45
+ | 42 | 0.3213 | 6 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8031 | 43.7707 | 0 |
46
+ | 43 | 0.3249 | 7 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9278 | 45.9357 | 0 |
47
+ | 44 | 0.289 | 8 | 0 | wikitext | comb_16 | opt | nlr_t_cos | 671.453 | 861.929 | 9.2263 | 47.9017 | 0 |
48
+ | 45 | 0.3253 | 7 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9308 | 48.0204 | 0 |
49
+ | 46 | 0.3244 | 7 | 0 | wikitext | comb_17 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9408 | 48.1147 | 0 |
50
+ | 47 | 0.3185 | 5 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 8.0651 | 48.6291 | 0 |
51
+ | 48 | 0.3244 | 8 | 0 | wikitext | comb_17 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9163 | 48.7731 | 0 |
52
+ | 49 | 0.3093 | 4 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 8.1976 | 49.0334 | 0 |
53
+ | 50 | 0.3173 | 5 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 8.0241 | 49.0665 | 0 |
54
+ | 51 | 0.3245 | 8 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9176 | 49.3775 | 0 |
55
+ | 52 | 0.3255 | 7 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8984 | 49.6926 | 0 |
56
+ | 53 | 0.3223 | 6 | 0 | wikitext | comb_18 | opt | nlr_t_cos | 671.453 | 861.929 | 7.7916 | 49.805 | 0 |
57
+ | 54 | 0.3223 | 6 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 7.7974 | 50.0844 | 0 |
58
+ | 55 | 0.318 | 8 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9519 | 50.978 | 0 |
59
+ | 56 | 0.3186 | 5 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 8.0729 | 52.2013 | 0 |
60
+ | 57 | 0.3216 | 6 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8115 | 52.8322 | 0 |
61
+ | 58 | 0.324 | 8 | 0 | wikitext | comb_5 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9276 | 54.346 | 0 |
62
+ | 59 | 0.178 | 7 | 0 | wikitext | comb_9 | opt | nlr_t_cos | 671.453 | 861.929 | 19.7972 | 55.1486 | 0 |
63
+ | 60 | 0.3228 | 8 | 0 | wikitext | comb_1 | opt | nlr_t_cos | 671.453 | 861.929 | 7.9688 | 55.2379 | 0 |
64
+ | 61 | 0.3245 | 8 | 0 | wikitext | comb_11 | opt | nlr_t_cos | 671.453 | 861.929 | 7.8961 | 55.7749 | 0 |
65
+ | 62 | 0.3076 | 4 | 0 | wikitext | comb_8 | opt | nlr_t_cos | 671.453 | 861.929 | 8.3183 | 59.3249 | 0 |
66
+ | 63 | 0 | 4 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.1121 | 0.0719 | 0 |
67
+ | 64 | 0 | 5 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.1121 | 0.0719 | 0 |
68
+ | 65 | 0 | 6 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.1121 | 0.0719 | 0 |
69
+ | 66 | 0 | 7 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.1121 | 0.0719 | 0 |
70
+ | 67 | 0 | 8 | 0 | wikitext | comb_14 | opt | nlr_t_cos | 477.74 | 968.884 | 1.1121 | 0.0719 | 0 |
71
+ | 68 | 0.3244 | 32 | 0 | wikitext | base | opt | ---- | 1359.4 | 1402.99 | 7.9349 | 19.5205 | 0 |