Good
#3
by Banaxi-Tech - opened
For experimental first model, this is insane, ONLY 0.1? Your going to do very well
Thanks for the support, but this model fell short of our goals. We know we can do better.
Next, we will be developing a new custom architecture trained on 30B more tokens.
How many tokens did this use if you want to say it
It says it in the model card, about 70B tokens.
Idk what i was thinking
Banaxi-Tech changed discussion status to closed
@Banaxi-Tech , God gave you eyes to look.
Now, open your eyes and LOOK AT THE README
Woah, no need to be so angry.