S21 Dflash needs about 36.3 GB of memory at IQ1_M (a 35.6 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 3 quants, from 2.2 GB (BF16) to 37.2 GB (IQ2_XXS). Longer context needs more: at 32K tokens the KV cache alone adds about 0.8 GB.
S21 Dflash is a 118B-parameter Laguna model in GGUF format, quantized by tepirale. Its trained context length is 1024K tokens. The repository has 3 quant files, from 2.2 GB (BF16) to 37.2 GB (IQ2_XXS).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| BF16 | 2.2 GB | 2.9 GB | Fits | Get file → | |
| IQ1_M | 35.6 GB | 36.3 GB | Too big | Get file → | |
| IQ2_XXS | 37.2 GB | 37.9 GB | Too big | Get file → |
What GPU do you need for S21 Dflash?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 6 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 8 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 12 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 16 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 24 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 32 GB VRAM: the best quant that fits is BF16 (2.2 GB file, about 2.9 GB in use).
- 48 GB VRAM: the best quant that fits is IQ2_XXS (37.2 GB file, about 37.9 GB in use).
Which S21 Dflash quant should you choose?
IQ1_M (35.6 GB) is the middle option in this repository. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
If memory is tight, the smallest file is BF16 at 2.2 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about S21 Dflash
How much VRAM does S21 Dflash need?
About 36.3 GB at IQ1_M with an 8K context window. That is the 35.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run S21 Dflash on an 8 GB GPU?
Yes, with BF16: the 2.2 GB file needs about 2.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of S21 Dflash GGUF?
It depends on the quant: BF16 is 2.2 GB, IQ1_M is 35.6 GB, IQ2_XXS is 37.2 GB.
What context length does S21 Dflash support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.8 GB at 32K.
Where can I download S21 Dflash GGUF?
The files are on Hugging Face in the tepirale/Laguna-S-2.1-DFLASH-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does S21 Dflash use?
Hugging Face lists the license as openmdw-1.1. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.