RSVP para seu evento TensorFlow Everywhere hoje mesmo!

tf.OptimizerOptions

View source on GitHub

A ProtocolMessage

do_common_subexpression_elimination bool do_common_subexpression_elimination
do_constant_folding bool do_constant_folding
do_function_inlining bool do_function_inlining
global_jit_level GlobalJitLevel global_jit_level
max_folded_constant_in_bytes int64 max_folded_constant_in_bytes
opt_level Level opt_level

Class Variables

  • DEFAULT = 0
  • GlobalJitLevel
  • L0 = -1
  • L1 = 0
  • Level
  • OFF = -1
  • ON_1 = 1
  • ON_2 = 2