Re:DeepSeek-R1 with Dynamic 1.58-bit Quantization
by Der_Einzige on Tuesday January 28, 2025 @03:13PM
(#42857363)
Btw, min_p (the paper about the sampler) got accepted to ICLR! As 4th author it warms my heart to so it used so much in the wild.
[ Parent
| on HN ]
Re:DeepSeek-R1 with Dynamic 1.58-bit Quantization
by danielhanchen on Tuesday January 28, 2025 @03:45PM
(#42857725)
Oh hi!! Congratulations on ICLR!!! min_p = 0.1 and temp = 1.5 is my default goto settings!!