Finished
WorldRecklessreductions-before-pruning-again-2diff8.0+0.08
LLR: -2.37 (-2.25, 2.89) [0.00, 4.00]
Games: 14876 W: 3632 L: 3696 D: 7548
Ptnml(0-2): 72, 1852, 3654, 1788, 72
WorldRecklessguard-qs-tt-cutoffdiff8.0+0.08
LLR: -2.41 (-2.25, 2.89) [0.00, 4.00]
Games: 8660 W: 2076 L: 2162 D: 4422
Ptnml(0-2): 41, 1120, 2106, 1010, 53
WorldRecklesshistory-above-betadiff8.0+0.08
LLR: -2.90 (-2.25, 2.89) [0.00, 4.00]
Games: 8092 W: 1969 L: 2077 D: 4046
Ptnml(0-2): 56, 980, 2050, 936, 24
WorldRecklessconthist-3diff8.0+0.08
LLR: -2.93 (-2.25, 2.89) [0.00, 4.00]
Games: 10952 W: 2608 L: 2710 D: 5634
Ptnml(0-2): 51, 1419, 2619, 1355, 32
PeregrRecklessimproving_change3diff8.0+0.08
LLR: -2.26 (-2.25, 2.89) [0.00, 4.00]
Games: 19492 W: 4772 L: 4816 D: 9904
Ptnml(0-2): 86, 2356, 4899, 2326, 79
PeregrRecklessimproving_change2diff8.0+0.08
LLR: -2.37 (-2.25, 2.89) [0.00, 4.00]
Games: 12982 W: 3137 L: 3206 D: 6639
Ptnml(0-2): 72, 1546, 3327, 1471, 75
PeregrRecklessrfp8diff8.0+0.08
LLR: -2.34 (-2.25, 2.89) [0.00, 4.00]
Games: 9008 W: 2152 L: 2232 D: 4624
Ptnml(0-2): 45, 1130, 2223, 1072, 34
PeregrRecklessrfp7diff8.0+0.08
LLR: -2.48 (-2.25, 2.89) [0.00, 4.00]
Games: 13520 W: 3267 L: 3340 D: 6913
Ptnml(0-2): 69, 1660, 3368, 1601, 62
StyxdoRecklesss2_lr_ub_increasediff40.0+0.40
LLR: -2.28 (-2.25, 2.89) [0.00, 4.00]
Games: 24380 W: 5778 L: 5807 D: 12795
Ptnml(0-2): 35, 2930, 6289, 2901, 35
LTC
StyxdoRecklesss2_lr_ub_increasediff8.0+0.08
LLR: -2.32 (-2.25, 2.89) [0.00, 4.00]
Games: 10394 W: 2529 L: 2606 D: 5259
Ptnml(0-2): 68, 1312, 2506, 1251, 60
Matching increase in S2 LR
PeregrRecklesslmr_stuff1diff8.0+0.08
LLR: 2.93 (-2.25, 2.89) [0.00, 4.00]
Games: 79640 W: 19446 L: 19071 D: 41123
Ptnml(0-2): 311, 9521, 19851, 9756, 381
WorldRecklesss1_upper_lr_bound_increasediff40.0+0.40
LLR: -2.28 (-2.25, 2.89) [0.00, 4.00]
Games: 26622 W: 6109 L: 6131 D: 14382
Ptnml(0-2): 37, 3075, 7095, 3081, 23
Increase S1 training LR Upper bound to initial_lr: 0.002 (up from 0.0015)
WorldRecklesspere-patchdiff8.0+0.08
LLR: -2.27 (-2.25, 2.89) [0.00, 4.00]
Games: 11644 W: 2793 L: 2862 D: 5989
Ptnml(0-2): 62, 1423, 2918, 1360, 59
WorldRecklesss1_upper_lr_bound_increasediff8.0+0.08
LLR: 3.08 (-2.25, 2.89) [0.00, 4.00]
Games: 11404 W: 2986 L: 2808 D: 5610
Ptnml(0-2): 60, 1291, 2824, 1465, 62
Increase S1 training LR Upper bound to initial_lr: 0.002 (up from 0.0015)
WorldRecklesss1_upper_lr_bound_increasediffN=25000
LLR: -2.26 (-2.25, 2.89) [0.00, 4.00]
Games: 5856 W: 1823 L: 1936 D: 2097
Ptnml(0-2): 189, 725, 1197, 644, 173
Increase S1 training LR Upper bound to initial_lr: 0.002 (up from 0.0015)
WorldRecklessfix-more-history-updatesdiff8.0+0.08
LLR: 3.08 (-2.25, 2.89) [-4.00, 0.00]
Games: 11382 W: 2780 L: 2675 D: 5927
Ptnml(0-2): 40, 1307, 2915, 1366, 63
as it turns out, we had another case where the history value got beyond the limit
WorldRecklessrfp-tt-pvdiff8.0+0.08
LLR: 2.91 (-2.25, 2.89) [0.00, 4.00]
Games: 18674 W: 4586 L: 4396 D: 9692
Ptnml(0-2): 68, 2200, 4644, 2324, 101
PeregrRecklessftard_ttqs2diff8.0+0.08
LLR: -2.27 (-2.25, 2.89) [0.00, 4.00]
Games: 17090 W: 4087 L: 4138 D: 8865
Ptnml(0-2): 68, 2041, 4377, 1992, 67
Take 2
WorldRecklessmore-static-historydiff8.0+0.08
LLR: -2.26 (-2.25, 2.89) [0.00, 4.00]
Games: 17156 W: 4094 L: 4144 D: 8918
Ptnml(0-2): 57, 2056, 4410, 1990, 65
WorldRecklessremove-null-move-historydiff8.0+0.08
LLR: 2.90 (-2.25, 2.89) [-4.00, 0.00]
Games: 15310 W: 3737 L: 3650 D: 7923
Ptnml(0-2): 61, 1847, 3771, 1896, 80
WorldRecklesspcm-bonus-5diff8.0+0.08
LLR: -1.93 (-2.25, 2.89) [0.00, 4.00]
Games: 7680 W: 1835 L: 1899 D: 3946
Ptnml(0-2): 34, 945, 1936, 901, 24
WorldRecklesspcm-bonus-5diff8.0+0.08
LLR: -2.37 (-2.25, 2.89) [0.00, 4.00]
Games: 7106 W: 1675 L: 1762 D: 3669
Ptnml(0-2): 36, 896, 1772, 817, 32
StyxdoRecklesshalf_batchsize_double_epochdiff40.0+0.40
LLR: -1.64 (-2.25, 2.89) [0.00, 4.00]
Games: 3156 W: 725 L: 786 D: 1645
Ptnml(0-2): 2, 403, 828, 344, 1
Extending the experiment to S1 too. Half the batchsize as v18, double the epochs.
WorldRecklessfix-factorizer-updatediff8.0+0.08
LLR: 2.91 (-2.25, 2.89) [-4.00, 0.00]
Games: 84500 W: 20323 L: 20446 D: 43731
Ptnml(0-2): 377, 10309, 20963, 10262, 339
StyxdoRecklesstest_s2_400_epochdiff40.0+0.40
LLR: -2.26 (-2.25, 2.89) [0.00, 4.00]
Games: 5554 W: 1234 L: 1316 D: 3004
Ptnml(0-2): 9, 695, 1449, 617, 7
Test changing to smaller batches but more epochs to match.