base_model: nvidia/Llama-3_1-Nemotron-51B-Instruct
language:
- en
library_name: transformers
license: other
license_link: >-
https://developer.download.nvidia.com/licenses/nvidia-open-model-license-agreement-june-2024.pdf
license_name: nvidia-open-model-license
quantized_by: mradermacher
tags:
- nvidia
- llama-3
- pytorch
About
weighted/imatrix quants of https://huggingface.co/nvidia/Llama-3_1-Nemotron-51B-Instruct
static quants are available at https://huggingface.co/mradermacher/Llama-3_1-Nemotron-51B-Instruct-GGUF
Usage
If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files.
Provided Quants
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
| Link | Type | Size/GB | Notes |
|---|---|---|---|
| GGUF | i1-IQ1_S | 11.5 | for the desperate |
| GGUF | i1-IQ1_M | 12.5 | mostly desperate |
| GGUF | i1-IQ2_XXS | 14.1 | |
| GGUF | i1-IQ2_XS | 15.6 | |
| GGUF | i1-IQ2_S | 16.6 | |
| GGUF | i1-IQ2_M | 17.9 | |
| GGUF | i1-Q2_K_S | 18.0 | very low quality |
| GGUF | i1-Q2_K | 19.5 | IQ3_XXS probably better |
| GGUF | i1-IQ3_XXS | 20.3 | lower quality |
| GGUF | i1-IQ3_XS | 21.7 | |
| GGUF | i1-IQ3_S | 22.8 | beats Q3_K* |
| GGUF | i1-Q3_K_S | 22.8 | IQ3_XS probably better |
| GGUF | i1-IQ3_M | 23.6 | |
| GGUF | i1-Q3_K_M | 25.3 | IQ3_S probably better |
| GGUF | i1-Q3_K_L | 27.4 | IQ3_M probably better |
| GGUF | i1-IQ4_XS | 27.8 | |
| GGUF | i1-Q4_0 | 29.4 | fast, low quality |
| GGUF | i1-Q4_K_S | 29.6 | optimal size/speed/quality |
| GGUF | i1-Q4_K_M | 31.1 | fast, recommended |
| GGUF | i1-Q4_1 | 32.5 | |
| GGUF | i1-Q5_K_S | 35.7 | |
| GGUF | i1-Q5_K_M | 36.6 | |
| GGUF | i1-Q6_K | 42.4 | practically like static Q6_K |
Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):
And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
FAQ / Model Request
See https://huggingface.co/mradermacher/model_requests for some answers to questions you might have and/or if you want some other model quantized.
Thanks
I thank my company, nethype GmbH, for letting me use its servers and providing upgrades to my workstation to enable this work in my free time. Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.