index
:
ben/bitsandbytes.git
main
fork of github.com/TimDettmers/bitsandbytes, trying to package it via nix
ben
summary
refs
log
tree
commit
diff
log msg
author
committer
range
path:
root
/
tests
Age
Commit message (
Expand
)
Author
2022-11-07
Fixed bug in cpu quant; faster GPU dequant.
Tim Dettmers
2022-11-06
Fixed blockwise test and logic.
Tim Dettmers
2022-11-06
Added blocksizes 2048, 1024, and 512 to blockwise quant.
Tim Dettmers
2022-11-06
Added k<256 quantile estimate.
Tim Dettmers
2022-11-06
Added k-bit fp8 map.
Tim Dettmers
2022-11-06
Added k-bit linear quantization.
Tim Dettmers
2022-11-03
Added FP8 quantization map.
Tim Dettmers
2022-10-25
Added CUDA SETUP instruction generator.
Tim Dettmers
2022-10-24
Isolated CUDASetup logging; all tests green.
Tim Dettmers
2022-09-20
try fp32
justheuristic
2022-09-20
set threshold
Tim Dettmers
2022-09-20
review
justheuristic
2022-09-18
cast device
justheuristic
2022-09-18
cast device
justheuristic
2022-09-18
cast device
justheuristic
2022-09-18
cast device
justheuristic
2022-09-18
cast device
justheuristic
2022-09-18
cast before allclose
justheuristic
2022-09-18
cast before allclose
justheuristic
2022-09-18
debug
justheuristic
2022-09-18
pre-cast
justheuristic
2022-09-18
debug
justheuristic
2022-09-18
run backward
justheuristic
2022-09-18
add memory efficient backward
justheuristic
2022-09-18
bump threshold to 0.21
justheuristic
2022-09-18
more lenient threshold
justheuristic
2022-09-18
cast properly
justheuristic
2022-09-17
un-fuse bias
justheuristic
2022-09-17
copypaste tolerances
justheuristic
2022-09-17
clearer assertions
justheuristic
2022-09-17
check dtypes first
justheuristic
2022-09-17
check dtypes first
justheuristic
2022-09-17
cast to half before double_quant
justheuristic
2022-09-17
test_bf16
justheuristic
2022-09-13
Fixed cpu blockwise quantization for small input tensors.
Tim Dettmers
2022-09-11
Fixed 2^31 max size issue for cpu blockwise quant.
Tim Dettmers
2022-08-23
Some initial code. Needs to be tested.
Tim Dettmers
2022-08-17
Fixed bug in Linear8bitLt, when the bias is None.
Tim Dettmers
2022-08-16
Added fused bias to matmullt.
Tim Dettmers
2022-08-16
Added fused bias in dequant_mm.
Tim Dettmers
2022-08-16
Removed storage() from get_ptr; added boilerplate for bias dequant_mm.
Tim Dettmers
2022-08-05
Added the case that all env variables are empty (CUDA docker).
Tim Dettmers
2022-08-04
Fixed bugs in cuda setup.
Tim Dettmers
2022-08-04
Merge branch 'debug' into cuda-bin-switch-and-cli
Tim Dettmers
2022-08-04
Merge branch 'extract_outliers' into debug
Tim Dettmers
2022-08-03
Added fixes for the case that matmullt dim A is zero, e.g. [0, 768].
Tim Dettmers
2022-08-02
factored cuda_setup.main out into smaller modules and functions
Titus von Koeller
2022-08-01
Added some more docs and comments.
Tim Dettmers
2022-08-01
Added full env variable search; CONDA_PREFIX priority.
Tim Dettmers
2022-08-01
reran black with linelength 80 for greater readability
Titus von Koeller
[next]