---
license: gemma
language:
- en
base_model:
- google/gemma-3-270m
tags:
- text-generation-inference
- transformers
- title-generation
- chat-summarization
- openwebui
- sft
- trl
- gemma3
pipeline_tag: text-generation
datasets:
- gghfez/openwebui_title_generation
---

# OpenWebUI Title Generator - Gemma 270M

A fine-tuned Gemma 270M model specialized for generating concise, emoji-enhanced titles for chat conversations.
Use this to avoid wasting API credits or wiping out the llama.cpp KV cache after the first message.

## Model Description

This model was fine-tuned on 8,900 examples of chat conversation title generation tasks. It takes chat history as input and outputs a JSON-formatted title with 3-5 words and relevant emojis.

## Intended Use

- Generate titles for chat conversations in OpenWebUI
- Summarize conversation themes quickly
- Create engaging, emoji-enhanced titles


## Performance

- Fast inference suitable for real-time title generation
- Lightweight at 270M parameters
- Consistent JSON output format


## Usage

Set this as the "Task Model" in OpenWebUI.

Settings -> Admin -> Interface

## Prompt format

The default OpenWebUI prompt format:

```
### Task:
Generate a concise, 3-5 word title with an emoji summarizing the chat history.
### Guidelines:
- The title should clearly represent the main theme or subject of the conversation.
- Use emojis that enhance understanding of the topic, but avoid quotation marks or special formatting.
- Write the title in the chat's primary language; default to English if multilingual.
- Prioritize accuracy over excessive creativity; keep it clear and simple.
### Output:
JSON format: { "title": "your concise title here" }
### Examples:
- { "title": "📉 Stock Market Trends" },
- { "title": "🍪 Perfect Chocolate Chip Recipe" },
- { "title": "Evolution of Music Streaming" },
- { "title": "Remote Work Productivity Tips" },
- { "title": "Artificial Intelligence in Healthcare" },
- { "title": "🎮 Video Game Development Insights" }
### Chat History:
<chat_history>
USER: hi
ASSISTANT: Hello! How can I assist you today? 😊
USER: who are you?
ASSISTANT: I’m Gemma, a large language model created by the Gemma team at Google DeepMind. I’m an open-weights model, which means I’m widely available for public use!
</chat_history>
```

Expected response:

```
'{"title": "👋 Simple Greeting"}'
```

## Limitations 

- Optimized specifically for title generation in OpenWebUI
- May not perform well in other chat interfaces
- Emojis sometimes not placed consistently or may be omitted entirely
- Only tested at BF16