xwinlm:70b-v0.1-q6_K

xwinlm:70b-v0.1-q6_K

101.7K Downloads Updated 1 year ago

Conversational model based on Llama 2 that performs competitively on various benchmarks.

7b 13b

Updated 1 year ago

1 year ago

003fda23145c · 57GB ·

archllama

·

parameters69B

·

quantizationQ6_K

57GB

{{ .System }} User: {{ .Prompt }} Assistant:

45B

{ "num_ctx": 4096, "num_gqa": 8, "stop": [ "User:", "Assistant:" ] }

58B

Readme

Xwin-LM is a model based on Llama 2 using various techniques to improve its quality.

Reference