asedmammad/
lhk-dpo:v1-q5_K_M

23 1 year ago

DPO finetuned model from FusionNet_7Bx2_MoE_14B

1 year ago

2aef458b7638 · 9.1GB ·

model
llama
·
12.9B
·
Q5_K_M
params
{ "stop": [ "[INST]", "[/INST]" ] }
template
[INST] {{ .System }} {{ .Prompt }} [/INST]

Readme

LHK_DPO_v1 is trained via Direct Preference Optimization(DPO) from https://huggingface.co/TomGrc/FusionNet_7Bx2_MoE_14B.

Original model is from https://huggingface.co/HanNayeoniee/LHK_DPO_v1