sorokin needs about 36.2 GB of memory at Q5_K_M (a 25.0 GB file) with an 8K context window, so it fits on a 48 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 42.9 GB.
sorokin is a 35B-parameter Sorokin model in GGUF format, quantized by man4j. Its trained context length is 128K tokens. The repository has 1 quant file, from 25.0 GB (Q5_K_M) to 25.0 GB (Q5_K_M).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q5_K_M | 25.0 GB | 36.2 GB | Too big | Get file → |
What GPU do you need for sorokin?
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: no quant fits at an 8K context.
- 32 GB VRAM: no quant fits at an 8K context.
- 48 GB VRAM: the best quant that fits is Q5_K_M (25.0 GB file, about 36.2 GB in use).
Which sorokin quant should you choose?
Q5_K_M (25.0 GB) is the middle option in this repository. Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about sorokin
How much VRAM does sorokin need?
About 36.2 GB at Q5_K_M with an 8K context window. That is the 25.0 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 sorokin on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 36.2 GB, so you would need a larger GPU or CPU offloading.
What is the file size of sorokin GGUF?
It depends on the quant: Q5_K_M is 25.0 GB.
What context length does sorokin support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 10.7 GB at 8K and 42.9 GB at 32K.
Where can I download sorokin GGUF?
The files are on Hugging Face in the man4j/sorokin_35b repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.