Skip to content

Peach-9B-8k-Roleplay

Run roleplay dialogue with a fine-tuned 9B model

Visit official site
Category AI chatbot

What Peach-9B-8k-Roleplay is

The model is described as a large language model fine-tuned for roleplay dialogue and based on 01-ai/Yi-1.5-9B. The model was trained with data synthesis on more than 100K conversations, and the source notes it may compare well with language models below 34B parameters.

What you can do with it

  • Generate dialogue text that fits a character setting
  • Hold roleplay conversations with the model
  • Use Chinese or English in dialogue
  • Adjust temperature and top_p to shape variety and coherence

Getting started

  • Import the needed libraries and modules, such as torch and transformers
  • Load the model and tokenizer with AutoTokenizer and AutoModelForCausalLM
  • Prepare dialogue messages with roles and content and tokenize them
  • Set generation parameters such as temperature and top_p

Model notes

  • Fine-tuning starts from 01-ai/Yi-1.5-9B
  • Training uses data synthesis on more than 100K conversations
  • Training data supports a maximum dialogue length of 8k
  • The source describes the parameter size as small

Pros & cons

✓ What we like

  • Fine-tuning is based on a large set of dialogue data
  • Training data supports conversations up to 8k in length

! What to watch out for

  • The model is built on 01-ai/Yi-1.5-9B
  • Generation is shaped by temperature and top_p settings

FAQ

What is the base model?

The material names 01-ai/Yi-1.5-9B as the base.

How much dialogue data was used?

The description states training used data synthesis on more than 100K conversations.

Last reviewed: 2026-11-15

More AI chatbot tools

View all →

How we review