Sulphur 2 base needs about 16.5 GB of memory at Q4_K_M (a 14.3 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 11 quants, from 10.3 GB (Q3_K_S) to 42.0 GB (BF16).
Sulphur 2 base is a 21B-parameter Sulphur model in GGUF format, quantized by Aiwithus. The repository has 11 quant files, from 10.3 GB (Q3_K_S) to 42.0 GB (BF16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_S | 10.3 GB | 12.1 GB | Too big | Get file → | |
| Q3_K_M | 11.1 GB | 13.0 GB | Too big | Get file → | |
| Q4_0 | 13.0 GB | 15.0 GB | Too big | Get file → | |
| Q4_K_S | 13.2 GB | 15.3 GB | Too big | Get file → | |
| Q4_K_M | 14.3 GB | 16.5 GB | Too big | Get file → | |
| Q5_K_S | 15.0 GB | 17.3 GB | Too big | Get file → | |
| Q5_0 | 15.3 GB | 17.6 GB | Too big | Get file → | |
| Q5_K_M | 16.1 GB | 18.6 GB | Too big | Get file → | |
| Q6_K | 17.8 GB | 20.4 GB | Too big | Get file → | |
| Q8_0 | 22.8 GB | 26.0 GB | Too big | Get file → | |
| BF16 | 42.0 GB | 47.6 GB | Too big | Get file → |
What GPU do you need for Sulphur 2 base?
These figures add the file size, an fp16 KV cache at an 8K context window 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: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: the best quant that fits is Q4_K_S (13.2 GB file, about 15.3 GB in use).
- 24 GB VRAM: the best quant that fits is Q6_K (17.8 GB file, about 20.4 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (22.8 GB file, about 26.0 GB in use).
- 48 GB VRAM: the best quant that fits is BF16 (42.0 GB file, about 47.6 GB in use).
Which Sulphur 2 base quant should you choose?
Q4_K_M is the usual starting point for Sulphur 2 base: it balances file size and quality well. Here it is a 14.3 GB file.
The Q8_0 file is 22.8 GB, about 1.6 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q3_K_S at 10.3 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Sulphur 2 base
How much VRAM does Sulphur 2 base need?
About 16.5 GB at Q4_K_M with an 8K context window. That is the 14.3 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 Sulphur 2 base on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 12.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Sulphur 2 base GGUF?
It depends on the quant: Q3_K_S is 10.3 GB, Q3_K_M is 11.1 GB, Q4_0 is 13.0 GB, Q4_K_S is 13.2 GB, Q4_K_M is 14.3 GB, Q5_K_S is 15.0 GB and 5 more.
Where can I download Sulphur 2 base GGUF?
The files are on Hugging Face in the Aiwithus/Sulphur-2-base-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.