![](/static/66c60d9f/assets/icons/icon-96x96.png)
![](https://lemmy.ml/pictrs/image/a64z2tlDDD.png)
3·
1 month agoBERT and early versions of GPT were trained on copyright free datasets like Wikipedia and out of copyright books. Unsure if those would be big enough for the modern ChatGPT types
BERT and early versions of GPT were trained on copyright free datasets like Wikipedia and out of copyright books. Unsure if those would be big enough for the modern ChatGPT types
Use reader mode a lot so love to see more tweaks available
that’s the Facebook font tho
You’ll never want to eat one again
I pray for the day web browsers get rid of “gecko Mozilla like”
There is a JavaScript API to describe what features it supports
What’s up with that? Appreciate they’re permissive rather than copyright free as such