chess-moe-router / train.log
avewright's picture
train.log
cd949ed verified
Raw History Blame Contribute Delete
4.77 kB
[18:43:21] step 1/4000 | loss=1.6037 | val_exp_ce=1.5518 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.491 trn_cls=1.5519 val_cls_ce=1.5518 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.493 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.06 rend=0.48 rope=0.45 rmid=0.00 | sacc ope=0.90 mid=0.00 end=0.91 gen=0.00 puz=0.16 | 0 pos/s
[18:45:24] step 50/4000 | loss=0.8224 | val_exp_ce=0.6046 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.781 trn_cls=0.5999 val_cls_ce=0.6046 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.779 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.27 rend=0.24 rope=0.25 rmid=0.24 | sacc ope=0.86 mid=0.68 end=0.86 gen=0.00 puz=0.72 | 0 pos/s
[18:47:28] step 100/4000 | loss=0.7456 | val_exp_ce=0.5210 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.799 trn_cls=0.5285 val_cls_ce=0.5210 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.809 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.25 rend=0.26 rope=0.26 rmid=0.24 | sacc ope=0.89 mid=0.72 end=0.91 gen=0.00 puz=0.71 | 0 pos/s
[18:49:32] step 150/4000 | loss=0.7291 | val_exp_ce=0.5010 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.812 trn_cls=0.5024 val_cls_ce=0.5010 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.812 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.21 rend=0.28 rope=0.23 rmid=0.28 | sacc ope=0.84 mid=0.82 end=0.95 gen=0.00 puz=0.64 | 0 pos/s
[18:51:37] step 200/4000 | loss=0.7371 | val_exp_ce=0.4727 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.824 trn_cls=0.4765 val_cls_ce=0.4727 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.824 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.21 rend=0.26 rope=0.27 rmid=0.26 | sacc ope=0.92 mid=0.80 end=0.91 gen=0.00 puz=0.66 | 0 pos/s
[18:53:41] step 250/4000 | loss=0.7243 | val_exp_ce=0.4696 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.829 trn_cls=0.4677 val_cls_ce=0.4696 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.825 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.23 rend=0.26 rope=0.22 rmid=0.28 | sacc ope=0.82 mid=0.84 end=0.93 gen=0.00 puz=0.71 | 0 pos/s
[18:55:45] step 300/4000 | loss=0.6781 | val_exp_ce=0.4518 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.835 trn_cls=0.4493 val_cls_ce=0.4518 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.832 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.22 rend=0.27 rope=0.25 rmid=0.25 | sacc ope=0.90 mid=0.80 end=0.94 gen=0.00 puz=0.69 | 0 pos/s
[18:57:49] step 350/4000 | loss=0.6778 | val_exp_ce=0.4499 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.831 trn_cls=0.4449 val_cls_ce=0.4499 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.835 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.21 rend=0.27 rope=0.25 rmid=0.27 | sacc ope=0.90 mid=0.83 end=0.95 gen=0.00 puz=0.67 | 0 pos/s
[18:59:53] step 400/4000 | loss=0.6792 | val_exp_ce=0.4448 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.843 trn_cls=0.4367 val_cls_ce=0.4448 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.836 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.24 rend=0.24 rope=0.26 rmid=0.25 | sacc ope=0.92 mid=0.79 end=0.89 gen=0.00 puz=0.74 | 0 pos/s
[19:01:58] step 450/4000 | loss=0.6814 | val_exp_ce=0.4354 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.839 trn_cls=0.4326 val_cls_ce=0.4354 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.839 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.23 rend=0.26 rope=0.26 rmid=0.25 | sacc ope=0.91 mid=0.79 end=0.93 gen=0.00 puz=0.72 | 0 pos/s
[19:04:02] step 500/4000 | loss=0.6852 | val_exp_ce=0.4337 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.846 trn_cls=0.4235 val_cls_ce=0.4337 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.839 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.22 rend=0.26 rope=0.26 rmid=0.26 | sacc ope=0.91 mid=0.82 end=0.93 gen=0.00 puz=0.70 | 0 pos/s
[19:06:07] step 550/4000 | loss=0.6869 | val_exp_ce=0.4296 val_route_ce=0.0000 prior_ce=0.0000 oracle_ce=0.0000 trn_route_ce=0.0000 trn_acc=0.832 trn_cls=0.4353 val_cls_ce=0.4296 o_rt=0.000 m_rt=0.000 e_rt=0.000 best_acc=0.839 | inc=0.0000 puz=0.0000 end=0.0000 ope=0.0000 mid=0.0000 | rinc=0.00 rpuz=0.23 rend=0.26 rope=0.27 rmid=0.23 | sacc ope=0.94 mid=0.76 end=0.94 gen=0.00 puz=0.72 | 0 pos/s