test / gpu_training /eval /evaluation_results.json
tanglumy
initial commit of model weights
1a6c70c
{
"gsm8k": {
"fr_ponder": {
"accuracy": 0.0,
"efficiency": {
"mean_flops": 1368148992.0,
"median_flops": 1554416640.0,
"std_flops": 1617190973.1658711,
"min_flops": 26234880.0,
"max_flops": 8591923200.0,
"total_flops": 136814899200.0
},
"speedup": {
"mean_time": 3.323502082824707,
"median_time": 4.929964542388916,
"std_time": 3.102591789855366,
"min_time": 0.08603239059448242,
"max_time": 9.481685638427734,
"total_time": 332.3502082824707
},
"avg_flops": 1368148992.0,
"avg_time": 3.323502082824707,
"avg_steps": 104.3
},
"baseline": {
"accuracy": 0.25,
"efficiency": {
"mean_flops": 1259616826016.48,
"median_flops": 1300493380979.0,
"std_flops": 61378209282.38763,
"min_flops": 1050444988777.0,
"max_flops": 1300493380979.0,
"total_flops": 125961682601648.0
},
"speedup": {
"mean_time": 5.597262227535248,
"median_time": 5.318989992141724,
"std_time": 2.637104224359403,
"min_time": 0.9660005569458008,
"max_time": 9.206240177154541,
"total_time": 559.7262227535248
},
"avg_flops": 1259616826016.48,
"avg_time": 5.597262227535248
},
"alpha_sweep": {
"0.1": {
"accuracy": 0.25,
"efficiency": {
"mean_flops": 314904206503.56,
"median_flops": 325123345244.0,
"std_flops": 15344552320.43663,
"min_flops": 262611247194.0,
"max_flops": 325123345244.0,
"total_flops": 31490420650356.0
},
"speedup": {
"mean_time": 5.451106414794922,
"median_time": 5.010058760643005,
"std_time": 2.636969014137301,
"min_time": 1.0382819175720215,
"max_time": 9.177106380462646,
"total_time": 545.1106414794922
},
"avg_flops": 314904206503.56,
"avg_time": 5.451106414794922
}
}
}
}