82 lines
2.3 KiB
Markdown
82 lines
2.3 KiB
Markdown
---
|
|
base_model:
|
|
- theprint/TiTan-Gemma3-4B
|
|
library_name: gguf
|
|
pipeline_tag: text-generation
|
|
language: en
|
|
license: apache-2.0
|
|
tags:
|
|
- gguf
|
|
- quantized
|
|
- llama.cpp
|
|
- titan-gemma3-4b
|
|
model_type: llama
|
|
quantized_by: theprint
|
|
datasets:
|
|
- theprint/titles-n-tags-alpaca
|
|
---
|
|
# TiTan-Gemma3-4B - GGUF Quantized
|
|
|
|
Quantized GGUF versions of [TiTan-Gemma3-4B](https://huggingface.co/theprint/TiTan-Gemma3-4B) for use with llama.cpp and other GGUF-compatible inference engines.
|
|
|
|
## Original Model
|
|
|
|
- **Base model:** [google/gemma-3-4b-it](https://huggingface.co/google/gemma-3-4b-it)
|
|
- **Fine-tuned model:** [theprint/TiTan-Gemma3-4B](https://huggingface.co/theprint/TiTan-Gemma3-4B)
|
|
- **Quantized by:** theprint
|
|
|
|
## Available Quantizations
|
|
|
|
- `TiTan-Gemma3-4B-f16.gguf` (8688.3 MB) - 16-bit float (original precision, largest file)
|
|
- `TiTan-Gemma3-4B-q3_k_m.gguf` (2276.3 MB) - 3-bit quantization (medium quality)
|
|
- `TiTan-Gemma3-4B-q4_k_m.gguf` (2734.6 MB) - 4-bit quantization (medium, recommended for most use cases)
|
|
- `TiTan-Gemma3-4B-q5_k_m.gguf` (3138.7 MB) - 5-bit quantization (medium, good quality)
|
|
- `TiTan-Gemma3-4B-q6_k.gguf` (3568.1 MB) - 6-bit quantization (high quality)
|
|
- `TiTan-Gemma3-4B-q8_0.gguf` (4619.2 MB) - 8-bit quantization (very high quality)
|
|
|
|
## Usage
|
|
|
|
### With llama.cpp
|
|
|
|
```bash
|
|
# Download recommended quantization
|
|
wget https://huggingface.co/theprint/TiTan-Gemma3-4B-GGUF/resolve/main/TiTan-Gemma3-4B-q4_k_m.gguf
|
|
|
|
# Run inference
|
|
./llama.cpp/main -m TiTan-Gemma3-4B-q4_k_m.gguf \
|
|
-p "Your prompt here" \
|
|
-n 256 \
|
|
--temp 0.7 \
|
|
--top-p 0.9
|
|
```
|
|
|
|
### With other GGUF tools
|
|
|
|
These files are compatible with:
|
|
- [llama.cpp](https://github.com/ggerganov/llama.cpp)
|
|
- [Ollama](https://ollama.ai/) (import as custom model)
|
|
- [KoboldCpp](https://github.com/LostRuins/koboldcpp)
|
|
- [text-generation-webui](https://github.com/oobabooga/text-generation-webui)
|
|
|
|
## Quantization Info
|
|
|
|
**Recommended:** `q4_k_m` provides the best balance of size, speed, and quality for most use cases.
|
|
|
|
**For maximum quality:** Use `q8_0` or `f16`
|
|
**For maximum speed/smallest size:** Use `q3_k_m` or `q4_k_s`
|
|
|
|
## License
|
|
|
|
apache-2.0
|
|
|
|
## Citation
|
|
|
|
```bibtex
|
|
@misc{titan_gemma3_4b_gguf,
|
|
title={TiTan-Gemma3-4B GGUF Quantized Models},
|
|
author={theprint},
|
|
year={2025},
|
|
publisher={Hugging Face},
|
|
url={https://huggingface.co/theprint/TiTan-Gemma3-4B-GGUF}
|
|
}
|
|
``` |