base_model: ssmits/Llama-3.1-Nemotron-93B-Instruct-HF-late
language:
- en
library_name: transformers
quantized_by: mradermacher
tags:
- mergekit
- merge
About
weighted/imatrix quants of https://huggingface.co/ssmits/Llama-3.1-Nemotron-93B-Instruct-HF-late
static quants are available at https://huggingface.co/mradermacher/Llama-3.1-Nemotron-93B-Instruct-HF-late-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 | 19.9 | for the desperate |
| GGUF | i1-IQ1_M | 21.7 | mostly desperate |
| GGUF | i1-IQ2_XXS | 24.8 | |
| GGUF | i1-IQ2_XS | 27.5 | |
| GGUF | i1-IQ2_S | 28.9 | |
| GGUF | i1-IQ2_M | 31.4 | |
| GGUF | i1-Q2_K | 34.2 | IQ3_XXS probably better |
| GGUF | i1-IQ3_XXS | 35.7 | lower quality |
| GGUF | i1-IQ3_XS | 38.1 | |
| GGUF | i1-Q3_K_S | 40.0 | IQ3_XS probably better |
| GGUF | i1-IQ3_S | 40.1 | beats Q3_K* |
| GGUF | i1-IQ3_M | 41.5 | |
| GGUF | i1-Q3_K_M | 44.5 | IQ3_S probably better |
| GGUF | i1-Q3_K_L | 48.4 | IQ3_M probably better |
| GGUF | i1-IQ4_XS | 49.4 | |
| PART 1 PART 2 | i1-Q4_0 | 52.3 | fast, low quality |
| PART 1 PART 2 | i1-Q4_K_S | 52.5 | optimal size/speed/quality |
| PART 1 PART 2 | i1-Q4_K_M | 55.4 | fast, recommended |
| PART 1 PART 2 | i1-Q5_K_S | 63.5 | |
| PART 1 PART 2 | i1-Q5_K_M | 65.2 | |
| PART 1 PART 2 | i1-Q6_K | 75.5 | 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.