I hate that part also at the same time enjoy it.
SnooHesitations8849
joined 10 months ago
Yes. Only when I need debugging. Otherwise, checking after a few hours is not too bad. Sometimes I know the code is correct, I just launch it and forget about it. Enjoying a few hours of doing nothing is better for your mental health than staring at the monitor gaining nothing.
A cheap macbook and a PC with a 3090 make much more sense.
Shader cores and tensor core only make sense when they are in a gaming card. Shader core carries the work of shading, while tensor core carries the upscaling, especially in case of Ray Tracing. In terms of operands, shading units works with vector as the base data structure, while in tensor core it is matrix (4x4 matrix) In ML card like A100, the amount of shading unit:tensor units is much lower than the ratio in gaming card.
Buy any mac with 16+GB of RAM. You wont be anle to train any model on any of those. So forget it. I am having a Macbook Pro with M2 Max and 96GB of RAM. Never attempt to train any model on it because it is a waste of time trying to do so.
Buy any mac with 16+GB of RAM. You wont be anle to train any model on any of those. So forget it. I am having a Macbook Pro with M2 Max and 96GB of RAM. Never attempt to train any model on it because it is a waste of time trying to do so.
Get a GPU with more memory like 12-16 GB. Dont need the CPU with the most #core. And if possible get 64GB of memory. Do not need 17 inches monitor.
Lower the python veréionto 3.5/3.6 then install pytorch 1.3.1
H100 and A100 are best for training. H100 is optimized for lower precision (8/16 bits) and optimized for transformer. A100 is still very good but not that much. A100 is still very GPU-like. Wwhile H100 is a transformer-accelerator.
Using them for inference is not the best econ-friendly though.
I would sell the card and buy s 1080Ti to offload all the burden.