सार्वजनिक स्थैतिक अंतिम एनम RewriterConfig.MemOptType
प्रोटोबफ एनम tensorflow.RewriterConfig.MemOptType
स्थिरांक
int यहाँ | DEFAULT_MEM_OPT_VALUE | The default setting (SCHEDULING and SWAPPING HEURISTICS only) DEFAULT_MEM_OPT = 0; |
int यहाँ | HEURISTICS_VALUE | Use any combination of swapping and recomputation heuristics. |
int यहाँ | मैनुअल_वैल्यू | Driven by manual op-level annotations. |
int यहाँ | NO_MEM_OPT_VALUE | Disabled in the meta-optimizer. |
int यहाँ | पुनर्गणना_HEURISTICS_VALUE | Recomputation heuristics will recompute ops (such as Relu activation) during backprop instead of storing them, reducing peak memory usage. |
int यहाँ | SCHEDULING_HEURISTICS_VALUE | Scheduling will split big ops such as AddN and try to enforce a schedule of the new computations that decreases peak memory usage. |
int यहाँ | SWAPPING_HEURISTICS_VALUE | Swapping heuristic will move a tensor from the GPU to the CPU and move it back when needed to reduce peak memory usage. |
विरासत में मिली विधियाँ
एनम मान
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType DEFAULT_MEM_OPT
The default setting (SCHEDULING and SWAPPING HEURISTICS only)
DEFAULT_MEM_OPT = 0;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType HEURISTICS
Use any combination of swapping and recomputation heuristics.
HEURISTICS = 3;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType मैनुअल
Driven by manual op-level annotations.
MANUAL = 2;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType NO_MEM_OPT
Disabled in the meta-optimizer.
NO_MEM_OPT = 1;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType RECOMPUTATION_HEURISTICS
Recomputation heuristics will recompute ops (such as Relu activation) during backprop instead of storing them, reducing peak memory usage.
RECOMPUTATION_HEURISTICS = 5;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType SCHEDULING_HEURISTICS
Scheduling will split big ops such as AddN and try to enforce a schedule of the new computations that decreases peak memory usage.
SCHEDULING_HEURISTICS = 6;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType SWAPPING_HEURISTICS
Swapping heuristic will move a tensor from the GPU to the CPU and move it back when needed to reduce peak memory usage.
SWAPPING_HEURISTICS = 4;
सार्वजनिक स्थैतिक अंतिम RewriterConfig.MemOptType अज्ञात
स्थिरांक
सार्वजनिक स्थैतिक अंतिम पूर्णांक DEFAULT_MEM_OPT_VALUE
The default setting (SCHEDULING and SWAPPING HEURISTICS only)
DEFAULT_MEM_OPT = 0;
स्थिर मान: 0
सार्वजनिक स्थैतिक अंतिम int HEURISTICS_VALUE
Use any combination of swapping and recomputation heuristics.
HEURISTICS = 3;
स्थिर मान: 3
सार्वजनिक स्थैतिक अंतिम पूर्णांक MANUAL_VALUE
Driven by manual op-level annotations.
MANUAL = 2;
स्थिर मान: 2
सार्वजनिक स्थैतिक अंतिम पूर्णांक NO_MEM_OPT_VALUE
Disabled in the meta-optimizer.
NO_MEM_OPT = 1;
स्थिर मान: 1
सार्वजनिक स्थैतिक अंतिम पूर्णांक RECOMPUTATION_HEURISTICS_VALUE
Recomputation heuristics will recompute ops (such as Relu activation) during backprop instead of storing them, reducing peak memory usage.
RECOMPUTATION_HEURISTICS = 5;
स्थिर मान: 5
सार्वजनिक स्थैतिक अंतिम पूर्णांक SCHEDULING_HEURISTICS_VALUE
Scheduling will split big ops such as AddN and try to enforce a schedule of the new computations that decreases peak memory usage.
SCHEDULING_HEURISTICS = 6;
स्थिर मान: 6
सार्वजनिक स्थैतिक अंतिम पूर्णांक SWAPPING_HEURISTICS_VALUE
Swapping heuristic will move a tensor from the GPU to the CPU and move it back when needed to reduce peak memory usage.
SWAPPING_HEURISTICS = 4;
स्थिर मान: 4