---

license: other

base_model: nvidia/Llama-3.1-Nemotron-Nano-8B-v1

base_model_relation: quantized

library_name: gguf

pipeline_tag: text-generation

tags:

• gguf
• llama.cpp
• dashq
• quantized

---

!DASH-Q

Llama-3.1-Nemotron-Nano-8B-v1 - DASH-Q 2-bit GGUF

llama.cpp GGUF files of nvidia/Llama-3.1-Nemotron-Nano-8B-v1 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 |

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

| Llama-3.1-Nemotron-Nano-8B-v1-DASHQ-IQ2_XXS.gguf | IQ2_XXS | 2.51 GB | 2.50 |

| Llama-3.1-Nemotron-Nano-8B-v1-DASHQ-IQ2_XS.gguf | IQ2_XS | 2.84 GB | 2.83 |

| Llama-3.1-Nemotron-Nano-8B-v1-DASHQ-IQ2_M.gguf | IQ2_M | 3.01 GB | 3.00 |

| Llama-3.1-Nemotron-Nano-8B-v1-DASHQ-Q2_K_XL.gguf | Q2_K_XL | 3.26 GB | 3.25 |

Perplexity (lower is better)

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

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

| IQ2_XXS | llama.cpp IQ2_XXS (imatrix) | 2.40 GB | 40.78 | 54.95 |

| IQ2_XXS | unsloth UD-IQ2_XXS | 2.50 GB | 36.30 | 47.86 |

| IQ2_XXS | DASH-Q IQ2_XXS | 2.51 GB | 27.81 | 41.99 |

| IQ2_XS | llama.cpp IQ2_XS (imatrix) | 2.61 GB | 33.04 | 45.14 |

| IQ2_XS | DASH-Q IQ2_XS | 2.84 GB | 21.90 | 33.47 |

| IQ2_M | llama.cpp IQ2_M (imatrix) | 2.95 GB | 19.51 | 30.20 |

| IQ2_M | unsloth UD-IQ2_M | 3.00 GB | 23.33 | 32.71 |

| IQ2_M | DASH-Q IQ2_M | 3.01 GB | 21.45 | 32.31 |

| Q2_K_XL | llama.cpp Q2_K (imatrix) | 3.18 GB | 27.15 | 38.60 |

| Q2_K_XL | unsloth UD-Q2_K_XL | 3.39 GB | 21.67 | 31.31 |

| Q2_K_XL | DASH-Q Q2_K_XL | 3.26 GB | 20.31 | 31.03 |

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

Usage

llama-cli -m Llama-3.1-Nemotron-Nano-8B-v1-DASHQ-Q2_K_XL.gguf -ngl 99 -c 8192

License

Inherits the license of the base model (nvidia/Llama-3.1-Nemotron-Nano-8B-v1).