eas/
openchat:latest

89 2 years ago

q4_k_m quantization only. Now using 8k context size

2 years ago

421ffc18355b · 4.4GB ·

model
llama
·
7.24B
·
Q4_K_M
system
You are a helpful AI assistant.
params
{ "num_ctx": 8192, "stop": [ "<|end_of_turn|>", "<|end\\_of\\_turn|>",
template
{{- if .First }}{{ .System }} <|end_of_turn>{{- end}}GPT4 Correct User: {{ .Prompt }}<|end_of_turn|>

Readme

I’ve only uploaded the -q4_k_m quantization

2023.11.06 Updated modelfile with PARAMETER num_ctx 8192

Original Model on HuggingFace

Quantized by TheBloke