Uploaded August 2026 | Updated September 2026, 17 hours ago
I came across this one posted by 'evanollama' on Ollama, and thought it should be really fast at 50m ... and so it is - but of course, no it won't translate. But a fun model nonetheless to use for quickly creating some children story content.
It appears it was trained on content with limited vocabulary, such as for kids up to 3-4 years old if I got this right.
So it's not a classical Large language model, it's a small (and fast) LM.
Would be interesting to add translation capability for just 1 target language, don't you agree? If training the engine on this content and bilingual aligned translations, could it still be of similar speed in just that one target language?
I came across this one posted by 'evanollama' on Ollama, and thought it should be really fast at 50m ... and so it is - but of course, no it won't translate. But a fun model nonetheless to use for quickly creating some children story content.
It appears it was trained on content with limited vocabulary, such as for kids up to 3-4 years old if I got this right.
So it's not a classical Large language model, it's a small (and fast) LM.
Would be interesting to add translation capability for just 1 target language, don't you agree? If training the engine on this content and bilingual aligned translations, could it still be of similar speed in just that one target language?
