---
base_model: Intelligent-Internet/II-Search-CIR-4B
---

[EXL3](https://github.com/turboderp-org/exllamav3) quantization of [II-Search-CIR-4B](https://huggingface.co/Intelligent-Internet/II-Search-CIR-4B), 4 bits per weight.

### HumanEval (argmax)

| Model                                                                                        | Q4   | Q6   | Q8   | FP16 |
| -------------------------------------------------------------------------------------------- | ---- | ---- | ---- | ---- |
| [II-Search-CIR-4B-exl3-4bpw](https://huggingface.co/isogen/II-Search-CIR-4B-exl3-4bpw)       | 81.7 | 79.3 | 78.7 | 79.9 |
| [II-Search-CIR-4B-exl3-6bpw](https://huggingface.co/isogen/II-Search-CIR-4B-exl3-6bpw)       | 80.5 | 81.1 | 81.1 | 81.7 |
| [II-Search-CIR-4B-exl3-8bpw-h8](https://huggingface.co/isogen/II-Search-CIR-4B-exl3-8bpw-h8) | 83.5 | 83.5 | 82.3 | 82.9 |
| [Qwen3-4B-exl3-4bpw](https://huggingface.co/isogen/Qwen3-4B-exl3-4bpw)                       | 80.5 | 81.1 | 81.7 | 80.5 |
| [Qwen3-4B-exl3-6bpw](https://huggingface.co/isogen/Qwen3-4B-exl3-6bpw)                       | 80.5 | 85.4 | 86.0 | 86.0 |
| [Qwen3-4B-exl3-8bpw-h8](https://huggingface.co/isogen/Qwen3-4B-exl3-8bpw-h8)                 | 82.3 | 84.8 | 83.5 | 82.9 |