times_22b.txt

 iteration        2/       5 | consumed samples:         2560 | consumed tokens:      5242880 | elapsed time per iteration (ms): 103920.6 | learning rate: 6.000E-05 | global batch size:  1280 | lm loss: 1.091125E+01 | loss scale: 1073741824.0 | grad norm: 0.000 | num zeros: 0.0 | actual seqlen:  2048 | number of skipped iterations:   0 | number of nan iterations:   0 | samples per second: 12.317 | TFLOPs: 292.40 |
 iteration        3/       5 | consumed samples:         3840 | consumed tokens:      7864320 | elapsed time per iteration (ms): 103049.9 | learning rate: 6.000E-05 | global batch size:  1280 | lm loss: 1.091024E+01 | loss scale: 536870912.0 | grad norm: 0.000 | num zeros: 0.0 | actual seqlen:  2048 | number of skipped iterations:   0 | number of nan iterations:   0 | samples per second: 12.421 | TFLOPs: 294.87 |
 iteration        4/       5 | consumed samples:         5120 | consumed tokens:     10485760 | elapsed time per iteration (ms): 102926.5 | learning rate: 6.000E-05 | global batch size:  1280 | lm loss: 1.091245E+01 | loss scale: 268435456.0 | grad norm: 0.000 | num zeros: 0.0 | actual seqlen:  2048 | number of skipped iterations:   0 | number of nan iterations:   0 | samples per second: 12.436 | TFLOPs: 295.22 |
 iteration        5/       5 | consumed samples:         6400 | consumed tokens:     13107200 | elapsed time per iteration (ms): 102787.8 | learning rate: 6.000E-05 | global batch size:  1280 | lm loss: 1.091016E+01 | loss scale: 134217728.0 | grad norm: 0.000 | num zeros: 0.0 | actual seqlen:  2048 | number of skipped iterations:   0 | number of nan iterations:   0 | samples per second: 12.453 | TFLOPs: 295.62 |