llama3-gradient

This model extends LLama-3 8B's context length from 8k to over 1m tokens.

8B 70B

66.1K Pulls Updated 4 months ago

... /
params
577073ffcc6c · 110B
{ "num_keep": 24, "stop": [ "<|start_header_id|>", "<|end_header_id|>", "<|eot_id|>" ] }