starkat99/half-rs

Issues

• rkyv support?

  #100 opened 6 months ago by comath
  1

• "cannot find attribute `derive` in this scope" and others with no_std

  #99 opened 9 months ago by ggoraa
  1

• Requiring Rust 1.70.0 and changing backwards compatibility policy in a minor version bump seems a mistake

  #97 opened a year ago by ThatGeoGuy
  1

• Introducing a breaking change "policy" on a minor version is problematic

  #96 opened a year ago by Kromey
  1

• Version 2.3.0 - CI breaks

  #93 opened a year ago by ChristopherRabotin
  3

• Run tests for i686 targets in CircleCI

  #95 opened a year ago by starkat99
  0

• Add support for ARM hardware intrinsics

  #63 opened a year ago by Shnatsel
  14

• Enable x86 F16C hardware intrinsics on stable Rust

  #54 opened a year ago by starkat99
  4

• Implements dtoa

  #92 opened a year ago by zzau13
  1

• lossless `TryFrom<f{32,64}>` implementations for `f16` and `bf16`

  #90 opened a year ago by Robbepop
  0

• need impl num_traits::ApproxEq for half data type

  #85 opened a year ago by Xudong-Huang
  0

• How do you cast into f16?

  #69 opened a year ago by BA8F0D39
  2

• Implement rand_distr::SampleUniform for f16

  #82 opened a year ago by ViliamVadocz
  2

• Don't build with serde support and `no_std`

  #84 opened a year ago by nathanielsimard
  2

• build fails when selecting serde feature but not std feature

  #79 opened a year ago by tspiteri
  2

• Fall back on C implementation of operators

  #74 opened a year ago by VariantXYZ
  3

• Add vector arithmetic operations

  #77 opened a year ago by starkat99
  0

• Improve CI solution for other architectures

  #72 opened a year ago by starkat99
  2

• Use 8x SIMD conversions when appropriate

  #66 opened a year ago by starkat99
  2

• Implement simba crate integration behind a feature flag

  #71 opened a year ago by farnoy
  0

• Add support for AVX-512-BF16 hardware intrinsics

  #70 opened a year ago by starkat99
  0

• Add support for AVX-512-FP16 hardware intrinsics

  #62 opened a year ago by starkat99
  0

• Allow rounding mode selection?

  #68 opened a year ago by starkat99
  0

• Expose `[f16; 4]` API

  #67 opened a year ago by starkat99
  0

• f16 serialization

  #60 opened 2 years ago by NazarKostetskiy
  3

• Unexpected behavior on `to_f32()`

  #59 opened 2 years ago by dam5h
  0

• support total ordering

  #55 opened 2 years ago by joseluis
  1

• Release prompt

  #52 opened 2 years ago by kali
  2

• Use `to_bits`/`from_bits` on `f32`/`f64` once `const` stabilized

  #53 opened 2 years ago by starkat99
  0

• Could from_f32() etc. be const-fn?

  #9 opened 2 years ago by Boscop
  5

• Resolver = 2 causing build issues with 1.8.1

  #48 opened 3 years ago by LucioFranco
  17

• Deprecation Note Markdown not working

  #28 opened 3 years ago by johannesvollmer
  0

• Implement the Float trait from the num-traits crate.

  #40 opened 3 years ago by frjnn
  1

• Debug for f16 and bf16 should convert to f32, like Display

  #37 opened 3 years ago by charles-r-earp
  1

• Should f16 implement `Mul`/`Add` traits, and expose methods to take powers (`powi`/`powf`)?

  #35 opened 4 years ago by bzm3r
  0

• Preserve the sign of NaNs when converted to f32?

  #23 opened 5 years ago by shaobo-he
  0

• Inconsistent rounding when converting from f32

  #24 opened 5 years ago by keram88
  0

• Support approximate comparisons using the `approx` crate

  #19 opened 5 years ago by tylerreisinger
  1

• Mutable Slice Conversion

  #16 opened 6 years ago by johannesvollmer
  1

• Convert an u16 array to f16 array

  #14 opened 6 years ago by johannesvollmer
  4

• f16::as_bits is not named like f32::to_bits and f64::to_bits

  #12 opened 6 years ago by tspiteri
  0

• Not able to build using rust 1.26

  #11 opened 6 years ago by JoshuaBatty
  3

• to_f32() visiblity

  #7 opened 6 years ago by PSeitz
  2

• converting from f64 nan produces infinity if only lower mantissa bits are nonzero

  #5 opened 6 years ago by tspiteri
  0

• comparison of +0 and -0 gives wrong result

  #2 opened 6 years ago by tspiteri
  1

• NaN behaviour of is_sign_positive and is_sign_negative is not like for primitive types

  #3 opened 6 years ago by tspiteri
  0

• Suggestion: derive Serialize and Deserialize for f16

  #1 opened 7 years ago by Boscop
  1