Llama 65b needs about 61.3 GB of memory at Q4_K_M (a 39.3 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 22 quants, from 7.4 GB (Q4_0) to 69.4 GB (Q8_0).
Llama 65b is a 65.3B-parameter Llama model in GGUF format, quantized by RichardErkhov. Its trained context length is 2K tokens. The repository has 22 quant files, from 7.4 GB (Q4_0) to 69.4 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 7.4 GB | 13.3 GB | Too big | Get file → | |
| Q2_K | 24.1 GB | 30.0 GB | Too big | Get file → | |
| IQ3_XS | 26.6 GB | 32.5 GB | Too big | Get file → | |
| IQ3_S | 28.2 GB | 34.0 GB | Too big | Get file → | |
| Q3_K_S | 28.2 GB | 34.0 GB | Too big | Get file → | |
| IQ3_M | 29.8 GB | 35.7 GB | Too big | Get file → | |
| Q3_K | 31.6 GB | 37.5 GB | Too big | Get file → | |
| Q3_K_M | 31.6 GB | 37.5 GB | Too big | Get file → | |
| Q3_K_L | 34.6 GB | 40.5 GB | Too big | Get file → | |
| IQ4_XS | 35.0 GB | 40.9 GB | Too big | Get file → | |
| IQ4_NL | 37.0 GB | 42.9 GB | Too big | Get file → | |
| Q4_K_S | 37.1 GB | 42.9 GB | Too big | Get file → | |
| Q4_K | 39.3 GB | 45.2 GB | Too big | Get file → | |
| Q4_K_M | 39.3 GB | 45.2 GB | Too big | Get file → | |
| Q4_1 | 40.9 GB (2 parts) | 46.7 GB | Too big | Get file → | |
| Q6_K | 42.3 GB (2 parts) | 48.2 GB | Too big | Get file → | |
| Q5_0 | 44.9 GB (2 parts) | 50.8 GB | Too big | Get file → | |
| Q5_K_S | 44.9 GB (2 parts) | 50.8 GB | Too big | Get file → | |
| Q5_K | 46.2 GB (2 parts) | 52.1 GB | Too big | Get file → | |
| Q5_K_M | 46.2 GB (2 parts) | 52.1 GB | Too big | Get file → | |
| Q5_1 | 49.0 GB (2 parts) | 54.9 GB | Too big | Get file → | |
| Q8_0 | 69.4 GB (2 parts) | 75.2 GB | Too big | Get file → |
What GPU do you need for Llama 65b?
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: the best quant that fits is Q4_0 (7.4 GB file, about 29.4 GB in use).
- 48 GB VRAM: the best quant that fits is Q2_K (24.1 GB file, about 46.1 GB in use).
Which Llama 65b quant should you choose?
Q4_K_M is the usual starting point for Llama 65b: it balances file size and quality well. Here it is a 39.3 GB file.
The Q8_0 file is 69.4 GB, about 1.8 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 Q4_0 at 7.4 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
Frequently asked questions about Llama 65b
How much VRAM does Llama 65b need?
About 61.3 GB at Q4_K_M with an 8K context window. That is the 39.3 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 Llama 65b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 29.4 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Llama 65b GGUF?
It depends on the quant: Q4_0 is 7.4 GB, Q2_K is 24.1 GB, IQ3_XS is 26.6 GB, IQ3_S is 28.2 GB, Q3_K_S is 28.2 GB, IQ3_M is 29.8 GB and 16 more.
What context length does Llama 65b support?
The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 21.5 GB at 8K and 5.4 GB at 2K.
Where can I download Llama 65b GGUF?
The files are on Hugging Face in the RichardErkhov/huggyllama_-_llama-65b-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 Llama 65b use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.