Stockmark 2 100B Instruct Beta needs about 38.9 GB of memory at Q2_K (a 35.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 13 quants, from 20.5 GB (IQ1_S) to 46.3 GB (Q3_K_M).
Stockmark 2 100B Instruct Beta is a 96B-parameter Stockmark model in GGUF format, quantized by mmnga. Its trained context length is 4K tokens. The repository has 13 quant files, from 20.5 GB (IQ1_S) to 46.3 GB (Q3_K_M).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_S | 20.5 GB | 22.4 GB | Too big | Get file → | |
| IQ1_M | 22.4 GB | 24.4 GB | Too big | Get file → | |
| IQ2_XXS | 25.7 GB | 27.6 GB | Too big | Get file → | |
| IQ2_XS | 28.5 GB | 30.4 GB | Too big | Get file → | |
| IQ2_S | 29.9 GB | 31.9 GB | Too big | Get file → | |
| IQ2_M | 32.5 GB | 34.5 GB | Too big | Get file → | |
| Q2_K | 35.5 GB | 37.5 GB | Too big | Get file → | |
| IQ3_XXS | 37.1 GB | 39.0 GB | Too big | Get file → | |
| IQ3_XS | 39.5 GB | 41.5 GB | Too big | Get file → | |
| Q3_K_S | 41.6 GB | 43.6 GB | Too big | Get file → | |
| IQ3_S | 41.7 GB | 43.7 GB | Too big | Get file → | |
| IQ3_M | 43.1 GB | 45.0 GB | Too big | Get file → | |
| Q3_K_M | 46.3 GB | 48.3 GB | Too big | Get file → |
What GPU do you need for Stockmark 2 100B Instruct Beta?
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: 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: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is IQ1_S (20.5 GB file, about 23.9 GB in use).
- 32 GB VRAM: the best quant that fits is IQ2_XS (28.5 GB file, about 31.9 GB in use).
- 48 GB VRAM: the best quant that fits is IQ3_M (43.1 GB file, about 46.5 GB in use).
Which Stockmark 2 100B Instruct Beta quant should you choose?
Q2_K (35.5 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 IQ1_S at 20.5 GB, with the quality trade-offs that come with fewer bits. 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.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Stockmark 2 100B Instruct Beta
How much VRAM does Stockmark 2 100B Instruct Beta need?
About 38.9 GB at Q2_K with an 8K context window. That is the 35.5 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 Stockmark 2 100B Instruct Beta on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 23.9 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Stockmark 2 100B Instruct Beta GGUF?
It depends on the quant: IQ1_S is 20.5 GB, IQ1_M is 22.4 GB, IQ2_XXS is 25.7 GB, IQ2_XS is 28.5 GB, IQ2_S is 29.9 GB, IQ2_M is 32.5 GB and 7 more.
What context length does Stockmark 2 100B Instruct Beta support?
The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 2.9 GB at 8K and 1.4 GB at 4K.
Where can I download Stockmark 2 100B Instruct Beta GGUF?
The files are on Hugging Face in the mmnga/Stockmark-2-100B-Instruct-beta-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 Stockmark 2 100B Instruct Beta use?
Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.