Age | Commit message (Expand) | Author |
---|---|---|
2022-08-25 | memory efficient fp16 backward | dbaranchuk |
2022-08-24 | minor fixes | dbaranchuk |
2022-08-23 | minor fixes | dbaranchuk |
2022-08-23 | refactoring | dbaranchuk |
2022-08-23 | add memory efficient backward | dbaranchuk |
2022-08-16 | Added fused bias to matmullt. | Tim Dettmers |
2022-08-08 | Fixed prod Python < 3.7 compatibility in function.py. | Tim Dettmers |
2022-08-08 | Removed prod for Python <= 3.7 compatibility. | 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 |
2022-08-01 | reran black with linelength 80 for greater readability | Titus von Koeller |
2022-08-01 | ran black and isort for coherent code formatting | Titus von Koeller |
2022-07-27 | Fixed direct extraction masking. | Tim Dettmers |
2022-07-26 | Matmullt with direct outlier extraction for 8-bit inference. | Tim Dettmers |
2022-07-22 | Most tests passing. | Tim Dettmers |