### 🔀 [#191](https://github.com/ikawrakow/ik_llama.cpp/pull/191) - Add additional checks for iq1_s_r4 quantization

| **Author** | `ikawrakow` |
| :--- | :--- |
| **State** | ❌ **Closed** |
| **Created** | 2025-02-07 |
| **Updated** | 2025-02-07 |

---

#### Description

Something goes wrong when quantizing DeepSeek-R1 with `IQ1_S_R4` (see #185), so adding additional checks in the quantization.