Starling is a large language model trained by reinforcement learning from AI feedback focused on improving chatbot helpfulness.
7b
62.1K Pulls Updated 7 months ago
4b21bfc435b4 · 200B
{{ if .System }}GPT4 Correct System: {{ .System }}<|end_of_turn|>{{ end }}{{ if .Prompt }}GPT4 Correct User: {{ .Prompt }}<|end_of_turn|>{{ end }}GPT4 Correct Assistant: {{ .Response }}<|end_of_turn|>