labadainofficial/ labadain-gemma4:e3b-q4km

10 Downloads Updated 6 days ago

Labadain-Gemma4 is a Gemma 4 model fine-tuned on 32,197 multi-turn conversations in Tetun, an official language of Timor-Leste.

tools thinking

ollama run labadainofficial/labadain-gemma4:e3b-q4km

curl http://localhost:11434/api/chat \
  -d '{
    "model": "labadainofficial/labadain-gemma4:e3b-q4km",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='labadainofficial/labadain-gemma4:e3b-q4km',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'labadainofficial/labadain-gemma4:e3b-q4km',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 6 days ago

6 days ago

f741e7526d7f · 3.4GB

model

archgemma4

·

parameters4.65B

·

quantizationQ4_K_M

3.4GB

params

{ "num_ctx": 2048, "stop": [ "<turn|>" ] }

46B

Readme

Labadain-Gemma4

Labadain-Gemma4 is a Gemma 4 model fine-tuned on 32,197 compiled Tetun multi-turn conversations from diverse sources covering general knowledge, news, everyday dialogue, health, and other domains.

Model Details

Model family: Gemma 4
Base model: unsloth/gemma-4-E2B-it
Primary language: Tetun
Fine-tuning: QLoRA
Training data: 32,197 compiled multi-turn conversations
Quantization: Q4_K_M
Developer: Labadain

Training Data

Source	Number of examples
News articles	1,951 multi-turn conversations
Health	9,649 QA and multi-turn conversations
Education and language	465 multi-turn conversations
General conversations	20,132 multi-turn conversations
Final compiled dataset	32,197 multi-turn conversations

Usage

Run the model with Ollama:

ollama run labadainofficial/labadain-gemma4:e3b-q4km

To disable thinking:

ollama run labadainofficial/labadain-gemma4:e3b-q4km --think=false

Intended Uses

Tetun conversational AI
Tetun text generation
Research on multilingual and low-resource NLP

Limitations

The model has not yet been evaluated on downstream applications.

Hugging Face

Merged model:
https://huggingface.co/labadain/Labadain-Gemma-4-E3B-it-Tetun

GGUF:
https://huggingface.co/labadain/Labadain-Gemma-4-E3B-it-Tetun-GGUF

License

This model is a fine-tuned derivative of Gemma 4. Users should comply with the applicable Gemma 4 license and terms when using or redistributing the model.

About Labadain

Labadain is a research initiative advancing information retrieval (IR), natural language processing (NLP), and artificial intelligence (AI) for Tetun and other low-resource language settings.

Related Projects

# Labadain-Gemma4

Labadain-Gemma4 is a Gemma 4 model fine-tuned on 32,197 compiled Tetun multi-turn conversations from diverse sources covering general knowledge, news, everyday dialogue, health, and other domains.

## Model Details

- **Model family:** Gemma 4
- **Base model:** `unsloth/gemma-4-E2B-it`
- **Primary language:** Tetun
- **Fine-tuning:** QLoRA
- **Training data:** 32,197 compiled multi-turn conversations
- **Quantization:** Q4_K_M
- **Developer:** Labadain

## Training Data

| Source | Number of examples |
|---|---:|
| News articles | 1,951 multi-turn conversations |
| Health | 9,649 QA and multi-turn conversations |
| Education and language | 465 multi-turn conversations |
| General conversations | 20,132 multi-turn conversations |
| **Final compiled dataset** | **32,197 multi-turn conversations** |

## Usage

Run the model with Ollama:

```bash
ollama run labadainofficial/labadain-gemma4:e3b-q4km
```

To disable thinking:

```bash
ollama run labadainofficial/labadain-gemma4:e3b-q4km --think=false
```

## Intended Uses

- Tetun conversational AI
- Tetun text generation
- Research on multilingual and low-resource NLP

## Limitations

**The model has not yet been evaluated on downstream applications.**

## Hugging Face

**Merged model:**  
https://huggingface.co/labadain/Labadain-Gemma-4-E3B-it-Tetun

**GGUF:**  
https://huggingface.co/labadain/Labadain-Gemma-4-E3B-it-Tetun-GGUF

## License

This model is a fine-tuned derivative of Gemma 4. Users should comply with the applicable Gemma 4 license and terms when using or redistributing the model.

## About Labadain

Labadain is a research initiative advancing information retrieval (IR), natural language processing (NLP), and artificial intelligence (AI) for Tetun and other low-resource language settings.

## Related Projects

- [Labadain Chat — Conversational AI for the Tetun Language](https://www.labadain.com)
- [Labadain Search - A Search Engine for the Tetun Language](https://search.labadain.com)

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)