Fine-tuned google/gemma-2-9b-it on princeton-nlp/gemma2-ultrafeedback-armorm with the SimPO objective.

948 3 months ago

c827688e53c8 · 137B
<start_of_turn>user
{{ if .System }}{{ .System }} {{ end }}{{ .Prompt }}<end_of_turn>
<start_of_turn>model
{{ .Response }}<end_of_turn>