---
license: apache-2.0
base_model: rinna/qwq-bakeneko-32b
base_model_relation: quantized
language:
- ja
---

Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.2.8">turboderp's ExLlamaV2 v0.2.8</a> for quantization.

**[2.2bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/2.2bpw)**  
**[3.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/3.0bpw)**  
**[4.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/4.0bpw)**  
**[5.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/5.0bpw)**  
**[6.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/6.0bpw)**  
**[7.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/7.0bpw)**  
**[8.0bpw](https://huggingface.co/rioshiina/qwq-bakeneko-32b-exl2/tree/8.0bpw)**

## Calibration Dataset
[TFMC/imatrix-dataset-for-japanese-llm](https://huggingface.co/datasets/TFMC/imatrix-dataset-for-japanese-llm)

## qwq-bakeneko-32b-exl2
- Model creator: [rinna](https://huggingface.co/rinna)
- Original model: [qwq-bakeneko-32b](https://huggingface.co/rinna/qwq-bakeneko-32b)

## License
[The Apache License, Version 2.0](https://opensource.org/license/apache-2-0)