Skip to main content

Clashluke's group workspace

Timestamps visible
2023-05-17 11:28:11
349 | /optimizer:0//body_ctx:0/reversible:7/write:3/scale_norm_act_linear:26/weight1/ema:0/momentum_buffer            | (8, 4096, 1024) | 134,217,728
2023-05-17 11:28:11
350 | /optimizer:0//body_ctx:0/reversible:7/write:3/scale_norm_act_linear:26/weight1/ema:1/momentum_buffer            | (8, 4096, 1024) | 134,217,728
2023-05-17 11:28:11
Parameters: 1,569,361,920
2023-05-17 11:28:11
Buffers:    3,138,723,840
2023-05-17 11:33:01
[    129/8388608] Loss:  4.683 - Accuracy:    0.095 | LearningRate: 0.00001 | StepTime: 291.181218s - Rate:     114.3 Tokens/s
2023-05-17 11:33:41
[    257/8388608] Loss:  4.661 - Accuracy:    0.101 | LearningRate: 0.00003 | StepTime:  31.630414s - Rate:     200.4 Tokens/s
2023-05-17 11:34:15
[    385/8388608] Loss:  4.469 - Accuracy:    0.090 | LearningRate: 0.00004 | StepTime:  31.632321s - Rate:     270.9 Tokens/s
2023-05-17 11:34:51
[    513/8388608] Loss:  4.237 - Accuracy:    0.101 | LearningRate: 0.00005 | StepTime:  31.639036s - Rate:     328.3 Tokens/s
2023-05-17 11:35:27
[    641/8388608] Loss:  3.999 - Accuracy:    0.099 | LearningRate: 0.00006 | StepTime:  31.634830s - Rate:     377.2 Tokens/s
2023-05-17 11:36:01
[    769/8388608] Loss:  3.621 - Accuracy:    0.111 | LearningRate: 0.00007 | StepTime:  31.633724s - Rate:     418.9 Tokens/s
2023-05-17 11:36:35
[    897/8388608] Loss:  3.517 - Accuracy:    0.124 | LearningRate: 0.00009 | StepTime:  31.642241s - Rate:     454.6 Tokens/s
2023-05-17 11:37:11
[   1025/8388608] Loss:  3.473 - Accuracy:    0.145 | LearningRate: 0.00010 | StepTime:  31.647683s - Rate:     485.2 Tokens/s
2023-05-17 11:37:49
[   1153/8388608] Loss:  3.447 - Accuracy:    0.154 | LearningRate: 0.00010 | StepTime:  31.643772s - Rate:     511.3 Tokens/s