ONE - On-device Neural Engine
|
Functions | |
float | relative_diff (float lhs, float rhs) |
Get the difference between two float values as a relative value. | |
bool | epsilon_equal (float expected, float obtained, uint32_t tolerance=1) |
Verify that an obtained float value is equal to the expected float value by using FLT_EPSILON. | |
bool | absolute_epsilon_equal (float expected, float obtained, float tolerance=0.001) |
Verify that an obtained float value is equal to the expected float value by comparing absolute tolerance value. | |
|
inline |
Verify that an obtained float value is equal to the expected float value by comparing absolute tolerance value.
[in] | expected | An expected float value to be compared |
[in] | obtained | An obtained float value to be compared |
[in] | tolerance | A tolerance value |
true
if both values are equal, otherwise false
Definition at line 82 of file fp32.h.
|
inline |
Verify that an obtained float value is equal to the expected float value by using FLT_EPSILON.
[in] | expected | An expected float value to be compared |
[in] | obtained | An obtained float value to be compared |
[in] | tolerance | A tolerance value |
true
if both values are equal, otherwise false
Definition at line 60 of file fp32.h.
|
inline |
Get the difference between two float values as a relative value.
[in] | lhs | A float value to be compared |
[in] | rhs | A float value to be compared |
Definition at line 44 of file fp32.h.