undefined | Better HN

0 pointsrayboy19955d ago0 comments

Thanks!! I had disabled that previously while debugging, I can confirm this is helping accuracy from what I can tell so far. (And speed since the cache is preserved more often!)

0 comments

satvikpendem5d ago

Use the MTP models which 2x token generation speed, for example: https://unsloth.ai/docs/models/qwen3.6#mtp-guide

rayboy1995OP5d ago

Very interesting I'll have to check this out thank you. This is why I love HN.

j / k navigate · click thread line to collapse

0 comments

satvikpendem5d ago

Use the MTP models which 2x token generation speed, for example: https://unsloth.ai/docs/models/qwen3.6#mtp-guide

rayboy1995OP5d ago

Very interesting I'll have to check this out thank you. This is why I love HN.

j / k navigate · click thread line to collapse