llama-pro:8b-instruct-q6_K

57.5K 1 year ago

An expansion of Llama 2 that specializes in integrating both general language understanding and domain-specific knowledge, particularly in programming and mathematics.

1 year ago

7df5d721dbc0 · 6.9GB

model
llama
·
8.36B
·
Q6_K
params
{ "stop": [ "<|system|>", "<|user|>", "<|assistant|>" ] }
template
<|system|> {{ .System }} <|user|> {{ .Prompt }} <|assistant|>

Readme

LLaMa-Pro is a version of the original LLaMa model enhanced by the addition of transformer blocks by Tencent Applied Research Center (ARC). This model specializes in integrating both general language understanding and domain-specific knowledge, particularly in programming and mathematics.

References

GitHub

Hugging Face