CodeLlama 34b needs about 69.6 GB of memory at F16 (a 67.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 3 quants, from 35.9 GB (Q8_0) to 135.0 GB (F32).
CodeLlama 34b is a 33.7B-parameter Llama model in GGUF format, quantized by Userb1az. Its trained context length is 16K tokens. The repository has 3 quant files, from 35.9 GB (Q8_0) to 135.0 GB (F32).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 35.9 GB | 38.0 GB | Too big | Get file → | |
| F16 | 67.5 GB | 69.6 GB | Too big | Get file → | |
| F32 | 135.0 GB | 137.1 GB | Too big | Get file → |
What GPU do you need for CodeLlama 34b?
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 Q8_0 (35.9 GB file, about 38.0 GB in use).
Which CodeLlama 34b quant should you choose?
F16 (67.5 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
If memory is tight, the smallest file is Q8_0 at 35.9 GB, with the quality trade-offs that come with fewer bits. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about CodeLlama 34b
How much VRAM does CodeLlama 34b need?
About 69.6 GB at F16 with an 8K context window. That is the 67.5 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 CodeLlama 34b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 38.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of CodeLlama 34b GGUF?
It depends on the quant: Q8_0 is 35.9 GB, F16 is 67.5 GB, F32 is 135.0 GB.
What context length does CodeLlama 34b support?
The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 1.6 GB at 8K and 3.2 GB at 16K.
Where can I download CodeLlama 34b GGUF?
The files are on Hugging Face in the Userb1az/CodeLlama-34b-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 CodeLlama 34b use?
Hugging Face lists the license as llama2. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.