apple

/

DCLM-7B-8k

like 45

Follow

Apple 6.7k

Hey, thanks for releasing this model! I am curious why does the title say 8K, the description also says that it has been trained to extend context length from 2k -> 8k. But at the top of the model card it says "Context Length 2048", and config.json also says "seq_len": 2048.