---

license: cc-by-4.0

base_model: nvidia/OpenReasoning-Nemotron-32B

base_model_relation: quantized

library_name: gguf

pipeline_tag: text-generation

tags:

• gguf
• llama.cpp
• dashq
• quantized

---

!DASH-Q

OpenReasoning-Nemotron-32B - DASH-Q 2-bit GGUF

llama.cpp GGUF files of nvidia/OpenReasoning-Nemotron-32B quantized with

DASH-Q, at several 2-bit-class sizes. Every file uses only standard

llama.cpp tensor types (no tensor above 4 bits) and loads in any recent llama.cpp build.

| File | Type | Size | Bits / weight |

| --- | --- | --- | --- |

| OpenReasoning-Nemotron-32B-DASHQ-IQ2_XXS.gguf | IQ2_XXS | 9.25 GB | 2.26 |

| OpenReasoning-Nemotron-32B-DASHQ-IQ2_XS.gguf | IQ2_XS | 10.59 GB | 2.59 |

| OpenReasoning-Nemotron-32B-DASHQ-IQ2_M.gguf | IQ2_M | 11.41 GB | 2.79 |

| OpenReasoning-Nemotron-32B-DASHQ-Q2_K_XL.gguf | Q2_K_XL | 12.52 GB | 3.06 |

Perplexity (lower is better)

| Type | Model | Size | WikiText-2 | C4 |

| --- | --- | --- | --- | --- |

| IQ2_XXS | llama.cpp IQ2_XXS (imatrix) | 9.03 GB | 71.33 | 87.46 |

| IQ2_XXS | unsloth UD-IQ2_XXS | 9.22 GB | 54.38 | 65.25 |

| IQ2_XXS | DASH-Q IQ2_XXS | 9.25 GB | 39.63 | 72.43 |

| IQ2_XS | llama.cpp IQ2_XS (imatrix) | 9.96 GB | 44.73 | 57.68 |

| IQ2_XS | DASH-Q IQ2_XS | 10.59 GB | 18.30 | 42.95 |

| IQ2_M | llama.cpp IQ2_M (imatrix) | 11.26 GB | 20.30 | 45.95 |

| IQ2_M | unsloth UD-IQ2_M | 11.39 GB | 22.37 | 45.39 |

| IQ2_M | DASH-Q IQ2_M | 11.41 GB | 18.59 | 43.44 |

| Q2_K_XL | llama.cpp Q2_K (imatrix) | 12.31 GB | 23.00 | 50.32 |

| Q2_K_XL | unsloth UD-Q2_K_XL | 12.70 GB | 16.78 | 35.68 |

| Q2_K_XL | DASH-Q Q2_K_XL | 12.52 GB | 17.82 | 41.29 |

llama-perplexity, context 2048; WikiText-2 test, C4 validation (256 x 2048 tokens).

Usage

llama-cli -m OpenReasoning-Nemotron-32B-DASHQ-Q2_K_XL.gguf -ngl 99 -c 8192

License

Inherits the license of the base model (nvidia/OpenReasoning-Nemotron-32B).