jshargo/ gemma-3N-finetune-4B:latest

38 1 year ago

A fine-tuned gemma3n 4b model trained on a wide variety of medical conversations

... /
template
1653b33bc82a · 378B
{{- range $i, $_ := .Messages }}
{{- $last := eq (len (slice $.Messages $i)) 1 }}
{{- if or (eq .Role "user") (eq .Role "system") }}<start_of_turn>user
{{ .Content }}<end_of_turn>
{{ if $last }}<start_of_turn>model
{{ end }}
{{- else if eq .Role "assistant" }}<start_of_turn>model
{{ .Content }}{{ if not $last }}<end_of_turn>
{{ end }}
{{- end }}
{{- end }}