3 ms·
That's why it feels like installing a ML repo is like sudoku. You install everything and at the last step you realize your neural net uses FlashAttention2 which
by visarga 2y ago
That's why it feels like installing a ML repo is like sudoku. You install everything and at the last step you realize your neural net uses FlashAttention2 which only works on NVIDIA compute version that is not deployed in your cloud VM and you need to start over from scratch.
- austinjp 2y agoThis describes the day I wasted on Monday before I gave up and wrote some damn deterministic code instead of using some damn AI.
- hskalin 2y agoSometimes I just change the version of the package in requirements to fit with others and pray that it works out (a few times it does)
- nicman23 2y agohonestly if the ml does not have a docker image - not compose no build an image- i do not even bother any more
- pjc50 2y agoSee the discussion on why sqlite insists on vendoring its build dependencies as far as possible and not using, say, CMake.
- anthk 2y agoGuix fixes that in the spot.