nohup: ignoring input
float division by zero
'int' object is not callable
'int' object is not callable
'int' object is not callable
device: cuda
./state_dict/roberta_wwm_pytorch_model.bin loaded!
data size is 205391
starting train.......
6000 /(3 /5 )
5 /(2 /5 -5 /11 )
3600 /2
loss is 4977.724117636681
6000 /(3 /5 )
5 /(5 /(5 +11 ) -2 /5 )
3600 /2
loss is 2840.0040685534477
6000 /(3 /5 )
5 /(5 /(5 +11 ) -2 /5 )
3600 /2
loss is 2410.8211997151375
validing..........
valid acc is 0.3895558223289316
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /(5 +11 ) -2 /5 )
3600 /2
loss is 2141.9377940297127
epoch is 0. loss is 12938.883185699582. spend time is 3652.494179725647
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 2 /5
3600 /2
loss is 1866.0435588508844
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 2 /5
3600 /2
loss is 1786.8222410529852
6000 /(1 +3 /5 )
5 /(5 /(11 -2 ) -5 /(11 -2 ) ) * 2 /(11 -2 )
3600 /2
loss is 1718.7221390157938
validing..........
valid acc is 0.5336134453781513
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 -(1 -2 /5 ) ) * 2 /5 -5 /11
3600 /2
loss is 1663.3558079972863
epoch is 1. loss is 7486.130717165768. spend time is 4022.7367458343506
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 2 /5
3600 /10
loss is 1493.030777156353
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 2 /5 -5 /(5 /11 -2 /5 ) * 2 /5
3600 /(1 /10 )
loss is 1452.8913628831506
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 5 /11 -5 /11
3600 /10
loss is 1440.0376018360257
validing..........
valid acc is 0.5960384153661464
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 2 /5
3600 /2
loss is 1404.6753779128194
epoch is 2. loss is 6184.924355499446. spend time is 4543.694051742554
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 -(1 -2 /5 ) ) * 2 /5 -5 /11
3600 /2
loss is 1272.219224371016
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 5 /11
3600 /2
loss is 1276.6901333630085
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * 5 /11
3600 /2
loss is 1277.3437282443047
validing..........
int too large to convert to float
valid acc is 0.6368547418967587
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * (1 -5 /11 )
3600 /2
loss is 1238.5479015409946
epoch is 3. loss is 5417.474597595632. spend time is 3349.9168536663055
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * (1 -5 /11 )
3600 /100
loss is 1129.589602611959
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * (1 -5 /11 )
3600 /2
loss is 1129.1439042836428
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * (1 -5 /11 )
6
loss is 1132.5003394186497
validing..........
valid acc is 0.6594637855142057
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) /(5 /11 -2 /5 )
3600 /1000
loss is 1146.307531364262
epoch is 4. loss is 4857.725408963859. spend time is 3288.392922639847
starting train.......
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /5 ) * 5 /(11 -5 )
3600 /10
loss is 1033.1504345834255
6000 /(1 +3 /5 )
5 /(5 /11 -2 /5 ) * (1 -5 /11 )
3600 /(3600 /2 )
loss is 1030.5476805120707
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(1 -2 /5 )
3600 /2
loss is 1032.0448204427958
validing..........
valid acc is 0.6800720288115246
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(1 -5 /11 )
3600 /100
loss is 1024.2290851548314
epoch is 5. loss is 4410.915047995746. spend time is 3317.6016352176666
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(5 /11 /(1
3600 /10
loss is 924.8407468907535
6000 /(1 +3 /5 )
5 /(5 /11 -(1 -2 /5 ) ) * (1 -5 /11 )
3600 /(3600 /10 )
loss is 941.9118327535689
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * (1 -5 /11 )
3600 /10
loss is 945.6323149614036
validing..........
valid acc is 0.6904761904761905
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 2 /5 /(1 -2 /5 )
3600 /100
loss is 944.1880606412888
epoch is 6. loss is 4024.077596012503. spend time is 3356.9466478824615
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 858.4414462819695
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
3600 /2
loss is 861.4646180532873
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) ) * (1 -5 /(11 -5 ) -2 /(5 -2 ) )
3600 /10
loss is 869.059286179021
validing..........
valid acc is 0.7018807523009204
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(1 -5 /11 )
3600 /10
loss is 869.5297578014433
epoch is 7. loss is 3696.2355847191066. spend time is 3288.027341604233
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * (1 -5 /11 )
3600 /2
loss is 779.8758879825473
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(1 -5 /11 )
3600 /2
loss is 790.2309758104384
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * 5 /11 /(1 -5 /11 )
3600 /10
loss is 802.4326879847795
validing..........
valid acc is 0.7146858743497398
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 808.7244394682348
epoch is 8. loss is 3403.673898259178. spend time is 3306.4831688404083
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 707.5494366008788
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * (1 -5 /11 )
60
loss is 724.9748096168041
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) /(1 -5 /11 )
60
loss is 738.5147444531322
validing..........
valid acc is 0.7184873949579832
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 743.0242428444326
epoch is 9. loss is 3126.740740245208. spend time is 3314.0832040309906
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
10
loss is 647.7412608359009
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 659.6942856907845
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 685.325991211459
validing..........
int too large to convert to float
valid acc is 0.7228891556622649
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) )
60
loss is 693.299962664023
epoch is 10. loss is 2880.3061905540526. spend time is 3268.9763791561127
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 595.4835678357631
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
3600 /90
loss is 613.1016807854176
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 627.2359251379967
validing..........
valid acc is 0.7216886754701881
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 638.1315435972065
epoch is 11. loss is 2648.79159244895. spend time is 3322.6155178546906
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) ) * (1 -5 /11 )
60
loss is 546.5388310085982
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 561.6939801946282
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 580.4377399748191
validing..........
valid acc is 0.7312925170068028
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 590.1024409867823
epoch is 12. loss is 2439.029150542803. spend time is 3268.309118270874
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 498.98468815535307
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) )
60
loss is 513.6433180961758
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 530.9659251142293
validing..........
valid acc is 0.7314925970388155
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 544.378740079701
epoch is 13. loss is 2241.207529697567. spend time is 3389.8482003211975
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 459.83615340292454
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 473.30991273373365
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 491.63859939482063
validing..........
valid acc is 0.7272909163665466
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 496.1610126327723
epoch is 14. loss is 2057.1158451242372. spend time is 3311.917845726013
starting train.......
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) )
60
loss is 423.7152771675028
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) )
60
loss is 438.73273359797895
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 442.39252999424934
validing..........
valid acc is 0.7302921168467387
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 457.1425550803542
epoch is 15. loss is 1893.237170251552. spend time is 3275.8674631118774
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 388.33942923881114
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 396.63631186354905
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 414.0540463635698
validing..........
valid acc is 0.7312925170068028
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 418.1898147892207
epoch is 16. loss is 1737.8617496620864. spend time is 3296.953103542328
starting train.......
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /5 )
60
loss is 352.9670407259837
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 366.27980756154284
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 377.44224847853184
validing..........
valid acc is 0.7376950780312125
./state_dict/bert_math_ques_model.bin saved!
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 391.7659137081355
epoch is 17. loss is 1597.471234726254. spend time is 3281.683824300766
starting train.......
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 322.4409961728379
6000 /(1 +3 /5 )
5 /(5 /11 /(1 -5 /11 ) -2 /5 /(1 -2 /5 ) )
60
loss is 335.94468710105866
6000 /(1 +3 /5 )
5 /(5 /(11 -5 ) -2 /(5 -2 ) )
60
loss is 353.5685621909797
validing..........