wimmmm/poolside-Laguna-S-2.1-GGUF overview
This is a GGUF quantization of Poolside's Laguna S 2.1 using an imatrix based on Bartowski's dataset. It was generated with llama.cpp 846e991e https://github.c…
Runs locally from ~334.9 MB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| laguna-s-2.1-IQ2_M.gguf | GGUF | IQ2_M | 35.75 GB | Download |
| laguna-s-2.1-IQ2_XXS.gguf | GGUF | IQ2_XXS | 28.71 GB | Download |
| laguna-s-2.1-IQ3_M.gguf | GGUF | IQ3_M | 47.96 GB | Download |
| laguna-s-2.1-IQ3_XXS.gguf | GGUF | IQ3_XXS | 42.14 GB | Download |
| laguna-s-2.1-IQ4_NL.gguf | GGUF | IQ4_NL | 61.80 GB | Download |
| laguna-s-2.1-IQ4_XS.gguf | GGUF | IQ4_XS | 58.39 GB | Download |
| laguna-s-2.1-MXFP4_MOE.gguf | GGUF | GGUF | 60.26 GB | Download |
| laguna-s-2.1-Q2_K.gguf | GGUF | Q2_K | 39.95 GB | Download |
| laguna-s-2.1-Q3_K_M.gguf | GGUF | Q3_K_M | 52.36 GB | Download |
| laguna-s-2.1-Q4_K_M.gguf | GGUF | Q4_K_M | 66.28 GB | Download |
| laguna-s-2.1-Q5_K_M.gguf | GGUF | Q5_K_M | 77.74 GB | Download |
| laguna-s-2.1-Q6_K.gguf | GGUF | Q6_K | 89.93 GB | Download |
| laguna-s-2.1-Q8_0.gguf | GGUF | Q8_0 | 116.44 GB | Download |
| laguna-s-2.1-imatrix.gguf | GGUF | GGUF | 334.9 MB | Download |
Model Details
Model README
---
license: openmdw-1.1
base_model:
- poolside/Laguna-S-2.1
---
This is a GGUF quantization of Poolside's Laguna S 2.1 using an imatrix based on Bartowski's dataset.
It was generated with llama.cpp 846e991e plus the changes from PR 25165, which at the time of building included commits up to 54f214a09b8
DFlash draft model is available at wimmmm/poolside-Laguna-S-2.1-DFlash-GGUF
To run with DFlash enabled, you can use the following command line (change quantizations as needed):
llama serve -hf wimmmm/poolside-Laguna-S-2.1-GGUF:Q4_K_M -hfd wimmmm/poolside-Laguna-S-2.1-DFlash-GGUF:Q4_0 --spec-type draft-dflash --spec-draft-n-max 15 -fa on --jinja --port 8000Have fun!
Run wimmmm/poolside-Laguna-S-2.1-GGUF with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models