Kenway

base, 11m params

a really tiny model

the details

size11M params
typetransformer
context now512 tokens
context goal4096 tokens
statebase (no instruct)
currently good atsimple text

context window

512 of 4096(target) 12.5%

a quick note this is a very small model. it can generate text, but don't expect miracles. instruct tuning is coming soon, which should make it a lot more usable. for now, just type anything and see what happens.

your prompt

generated content

0.00 tok/s

nothing yet. type something above to get started.