Age | Commit message (Collapse) | Author | |
---|---|---|---|
2022-12-21 | Log binary_path when CUDA setup failsmain | Ben Sima | |
This way we can debug the build paths if it fails. | |||
2022-11-20 | Added additional blocksizes: {64, 128, 256}. | Tim Dettmers | |
2022-11-19 | Fixed k-bit quantization maps. | Tim Dettmers | |
2022-11-07 | Fixed bug in cpu quant; faster GPU dequant. | Tim Dettmers | |
2022-11-06 | Added pre/post calls do quantize_blockwise. | 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 | |
2022-10-31 | Fixed CUDA setup bugs, including #81. | Tim Dettmers | |
2022-10-27 | Fixing bad error when GPU was not detected for #73. | Tim Dettmers | |
2022-10-26 | Replace seemingly incorrect use of CUDA_RUNTIME_LIB | Tom Aarsen | |
2022-10-26 | Resolve cases of CUDASetup.get_instance not being called when used | Tom Aarsen | |
2022-10-25 | Added CUDA SETUP instruction generator. | Tim Dettmers | |
2022-10-24 | Isolated CUDASetup logging; all tests green. | Tim Dettmers | |
2022-10-09 | Merge pull request #36 from tomaarsen/hotfix/os_error_name_too_long | Tim Dettmers | |
Fixes `OSError: File name too long` when environment variable is too long | |||
2022-10-09 | Merge pull request #37 from tomaarsen/hotfix/colab_just_cpu | Tim Dettmers | |
Perform check using implicit list length | |||
2022-09-20 | review | Tim Dettmers | |
2022-09-18 | debug | justheuristic | |
2022-09-18 | debug | justheuristic | |
2022-09-18 | debugprint | justheuristic | |
2022-09-18 | debugpritn | justheuristic | |
2022-09-18 | add memory efficient backward | justheuristic | |
2022-09-18 | reduce diff | justheuristic | |
2022-09-18 | rollback | justheuristic | |
2022-09-18 | addmm_ | justheuristic | |
2022-09-18 | cast edge case | justheuristic | |
2022-09-18 | cast edge case | justheuristic | |
2022-09-18 | cast properly | justheuristic | |
2022-09-18 | cast properly | justheuristic | |
2022-09-18 | cast properly | justheuristic | |
2022-09-18 | debug asset | justheuristic | |
2022-09-18 | matmul -1- addmm | justheuristic | |
2022-09-18 | safer cast | justheuristic | |
2022-09-18 | change typecast behavior | justheuristic | |
2022-09-18 | change typecast behavior | justheuristic | |
2022-09-18 | change typecast behavior | justheuristic | |
2022-09-18 | change typecast behavior | justheuristic | |
2022-09-18 | change typecast behavior | justheuristic | |
2022-09-18 | debug assert | justheuristic | |
2022-09-18 | debug assert | justheuristic | |
2022-09-17 | change order | justheuristic | |
2022-09-17 | change order | justheuristic | |
2022-09-17 | un-fuse bias | justheuristic | |
2022-09-17 | un-fuse bias | justheuristic | |
2022-09-17 | un-fuse bias | justheuristic | |
2022-09-17 | cast bias too | justheuristic | |
2022-09-17 | recast to fp16 | justheuristic | |