Skip to main content

Module numeric_translators

Module numeric_translators 

Source

Structsยง

BFloat16Translator
DoublePrecisionTranslator
E4M3Translator
E5M2Translator
HalfPrecisionTranslator
Posit8Translator
Posit16Translator
Posit32Translator
PositQuire8Translator
PositQuire16Translator
SignedFixedPointTranslator
SignedTranslator
SinglePrecisionTranslator
UnsignedFixedPointTranslator
UnsignedTranslator

Functionsยง

compute_signed_value ๐Ÿ”’
Computes the signed value string for a given BigUint and bit width.
decode_e4m3 ๐Ÿ”’
Decode u8 as 8-bit float with four exponent bits and three mantissa bits.
decode_e4m3_f64 ๐Ÿ”’
Decode u8 as 8-bit float with four exponent bits and three mantissa bits, returning f64.
decode_e5m2 ๐Ÿ”’
Decode u8 as 8-bit float with five exponent bits and two mantissa bits.
decode_e5m2_f64 ๐Ÿ”’
Decode u8 as 8-bit float with five exponent bits and two mantissa bits, returning f64.
format_fp8_value ๐Ÿ”’
Format an f64 value as a string for FP8 display.
shortest_float_representation ๐Ÿ”’
translate_numeric ๐Ÿ”’
If value is a biguint or consists only of 1 or 0, translates the value using biguint_translator. If value contains other values such as X, Z etc. the result is the corresponding ValueKind