Commits · unstable · bandicoot-lib / bandicoot-code

Browse files

Oct 31, 2025
- Merge branch 'fix_docs' into 'unstable' · d8661b69
  conradsnicta authored Oct 31, 2025
  d8661b69
- add missing opening and closing tags · d9280063
  conrad authored Oct 31, 2025
  
  d9280063
- Merge branch 'update_citation' into 'unstable' · 4b4b6446
  conradsnicta authored Oct 31, 2025
  4b4b6446
- update citation · 7dbf51cb
  conradsnicta authored Oct 31, 2025
  
  7dbf51cb
Oct 29, 2025
- Merge branch 'fp16' into 'unstable' · bbf247f3
  Ryan Curtin authored Oct 29, 2025
  bbf247f3
- Adjust tolerance. · 7997207b
  Ryan Curtin authored Oct 29, 2025
  
  7997207b
- Bump version and update documentation. · 1b74ddc0
  Ryan Curtin authored Oct 29, 2025
  
  1b74ddc0
Oct 28, 2025
- Bump to newer CUDA toolkit version. · 3ac48df0
  Ryan Curtin authored Oct 28, 2025
  
  3ac48df0
- Add missing overload of coot_is_fp(). · 9c677f8c
  Ryan Curtin authored Oct 28, 2025
  
  9c677f8c
- Fix compilation warnings. · ae3439bf
  Ryan Curtin authored Oct 28, 2025
  
  ae3439bf
- Use older CUDA functions for half/float/double conversions. · bd067760
  Ryan Curtin authored Oct 28, 2025
  
  bd067760
- Use manual casting with functions instead to support older CUDA toolkits. · 03b5089d
  Ryan Curtin authored Oct 28, 2025
  
  03b5089d
Oct 26, 2025
- Only include stdfloat for C++23 compilers. · 608b3d6c
  Ryan Curtin authored Oct 26, 2025
  
  608b3d6c
- Reduce probability of expr_cor_small fp16 test failing. · ca8dcdac
  Ryan Curtin authored Oct 26, 2025
  
  ca8dcdac
- Fix dot() computation for C++23 on OpenCL. · b80cbd43
  Ryan Curtin authored Oct 26, 2025
  
  b80cbd43
Oct 24, 2025
- Remove debugging output. · ee521e0f
  Ryan Curtin authored Oct 24, 2025
  
  ee521e0f
- Additional CUDA C++23 fixes. · 66004468
  Ryan Curtin authored Oct 24, 2025
  
  66004468
- Handle std::float16_t correctly for OpenCL. · ad4010f5
  Ryan Curtin authored Oct 24, 2025
  
  ad4010f5
- Update to work with std::float16_t if using C++23. · 6aefd395
  Ryan Curtin authored Oct 24, 2025
  
  6aefd395
Oct 23, 2025
- Fix tests to handle fp16_shim types more correctly. · 96a56cca
  Ryan Curtin authored Oct 23, 2025
  
  96a56cca
- Fix OpenCL backend implementations. · 88a587d8
  Ryan Curtin authored Oct 23, 2025
  
  88a587d8
- Fix fp16_shim handling. · 1ba06b0f
  Ryan Curtin authored Oct 23, 2025
  
  1ba06b0f
Oct 21, 2025
- Merge remote-tracking branch 'origin/unstable' into fp16 · 4c880223
  Ryan Curtin authored Oct 21, 2025
  
  4c880223
- Fixes for OpenCL backend. · 4d343bca
  Ryan Curtin authored Oct 21, 2025
  
  4d343bca
- Use h_kernels instead of fp16_kernels. · 8c3acf3c
  Ryan Curtin authored Oct 21, 2025
  
  8c3acf3c
- Fix placement in example config file. · d5458155
  Ryan Curtin authored Oct 21, 2025
  
  d5458155
- Merge branch 'replace' into 'unstable' · 30a661e3
  Ryan Curtin authored Oct 21, 2025
  30a661e3
Oct 20, 2025
- Add cl_type.hpp. · fae4893a
  Ryan Curtin authored Oct 20, 2025
  
  fae4893a
- Add working FP16 CUDA support and partial OpenCL support. · 73442a39
  Ryan Curtin authored Oct 20, 2025
  
  73442a39
Oct 17, 2025
- Update tolerances to reduce random test failures. · 9bcba773
  Ryan Curtin authored Oct 17, 2025
  
  9bcba773
Sep 30, 2025
- First attempt to adapt the rest of the tests to use low-precision FP types. · aaebd84a
  Ryan Curtin authored Sep 30, 2025
  
  aaebd84a
Sep 24, 2025
- Incremental commit, partially worked through the tests. · fd7a5bc5
  Ryan Curtin authored Sep 24, 2025
  
  fd7a5bc5
Sep 18, 2025
- Implement standalone replace(). · 520e5e87
  Ryan Curtin authored Sep 18, 2025
  
  520e5e87
- Merge branch 'lowprec-int-types' into 'unstable' · b70d11b8
  Ryan Curtin authored Sep 18, 2025
  b70d11b8
Sep 17, 2025
- Disable some tests that overflow. · d2b435de
  Ryan Curtin authored Sep 17, 2025
  
  d2b435de
- Update documentation. · 6c6a57dd
  Ryan Curtin authored Sep 17, 2025
  
  6c6a57dd
- Adapt OpenCL backend to u8/s8/u16/s16. · 14255672
  Ryan Curtin authored Sep 17, 2025
  
  14255672
- Update the rest of the tests so they work for low-precision on CUDA. · 2a5a0b96
  Ryan Curtin authored Sep 17, 2025
  
  2a5a0b96
Sep 16, 2025
- Further adaptation of tests. · 3a8bb3c0
  Ryan Curtin authored Sep 16, 2025
  
  3a8bb3c0
- Enable other types for randi generation on CUDA backend. · f77ef43d
  Ryan Curtin authored Sep 16, 2025
  
  f77ef43d