vitoplantamura

3 ms·

1. 1.▲
   Show HN: OnnxStream running TinyLlama and Mistral 7B, with CUDA support(github.com)

   2 points by vitoplantamura 3y ago|0 comments
2. 2.▲
   by vitoplantamura 3y ago

   By "0.5x slower" I mean "50% slower" :-)
3. 3.▲
   by vitoplantamura 3y ago

   hi, I'm the author. I have never questioned the clarity of that sentence, at least until today :-) By "0.5x" I mean "0.5 times or 0.5 multiplied by the reference time" where "reference time" is the inferen
4. 4.▲
   Show HN: BugChecker > SoftICE-like Kernel Debugger for Windows 11(github.com)

   6 points by vitoplantamura 4y ago|2 comments