llama.cpp needs about 8.6 GB of memory at Q4_0 (a 3.8 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 4 quants, from 3.8 GB (Q4_0) to 27.0 GB (F32).
llama.cpp is a 6.74B-parameter Llama model in GGUF format, quantized by Tien203. Its trained context length is 4K tokens. The repository has 4 quant files, from 3.8 GB (Q4_0) to 27.0 GB (F32).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 3.8 GB | 6.5 GB | Fits | Get file → | |
| Q4_0 | 3.8 GB | 6.5 GB | Fits | Get file → | |
| F16 | 13.5 GB | 16.1 GB | Too big | Get file → | |
| F32 | 27.0 GB | 29.6 GB | Too big | Get file → |
What GPU do you need for llama.cpp?
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: the best quant that fits is Q4_0 (3.8 GB file, about 8.6 GB in use).
- 16 GB VRAM: the best quant that fits is Q4_0 (3.8 GB file, about 8.6 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (13.5 GB file, about 18.3 GB in use).
- 32 GB VRAM: the best quant that fits is F32 (27.0 GB file, about 31.7 GB in use).
Which llama.cpp quant should you choose?
Q4_0 (3.8 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is Q4_0 at 3.8 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about llama.cpp
How much VRAM does llama.cpp need?
About 8.6 GB at Q4_0 with an 8K context window. That is the 3.8 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 llama.cpp on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 8.6 GB, so you would need a larger GPU or CPU offloading.
What is the file size of llama.cpp GGUF?
It depends on the quant: Q4_0 is 3.8 GB, Q4_0 is 3.8 GB, F16 is 13.5 GB, F32 is 27.0 GB.
What context length does llama.cpp support?
The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 4.3 GB at 8K and 2.1 GB at 4K.
Where can I download llama.cpp GGUF?
The files are on Hugging Face in the Tien203/llama.cpp repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.