koesn/
llama3-8b-instruct:Q4_0

1,217 1 year ago

Fixed num_ctx to 8192 and eos token. This Llama 3 8B Instruct model is ready to use for full model's 8k contexts window.

... /
params
15392c1ab95c · 129B
{
"num_ctx": 8192,
"num_keep": 24,
"stop": [
"<|start_header_id|>",
"<|end_header_id|>",
"\"<|eot_id|>\""
]
}