LL

llama v2 2bit

LlamaGGUF by ikawrakowUpdated 2023-12-06
Download on Hugging Face
Parameters13B
Context4K
Layers32
Architecturellama
Smallest quant2.3 GB
Quick answer

llama v2 2bit needs about 9.1 GB of memory at its main file (a 4.4 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 3 quants, from 2.3 GB to 22.9 GB.

llama v2 2bit is a 13B-parameter Llama model in GGUF format, quantized by ikawrakow. Its trained context length is 4K tokens. The repository has 3 quant files, from 2.3 GB to 22.9 GB.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—2.3 GB 4.9 GB
Fits Get file →
—4.4 GB 7.0 GB
Fits Get file →
—22.9 GB 25.6 GB
Too big Get file →

What GPU do you need for llama v2 2bit?

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: the best quant that fits is the main file (2.3 GB file, about 7.1 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (4.4 GB file, about 9.1 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (4.4 GB file, about 9.1 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (4.4 GB file, about 9.1 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (22.9 GB file, about 27.7 GB in use).

Which llama v2 2bit quant should you choose?

If memory is tight, the smallest file is the first one in the table at 2.3 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters13B
Layers32
Trained context4,096 tokens
KV cache (fp16)512 KB per token

Source

Quantized byikawrakow
FamilyLlama
Licenseapache-2.0
Downloads (HF)135
Last updated2023-12-06

Frequently asked questions about llama v2 2bit

How much VRAM does llama v2 2bit need?

About 9.1 GB at its main file with an 8K context window. That is the 4.4 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 v2 2bit on an 8 GB GPU?

Yes, with the smaller files: the 2.3 GB file needs about 7.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of llama v2 2bit GGUF?

It depends on the quant: file is 2.3 GB, file is 4.4 GB, file is 22.9 GB.

What context length does llama v2 2bit 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 v2 2bit GGUF?

The files are on Hugging Face in the ikawrakow/llama-v2-2bit-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 v2 2bit use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Similar models

Browse all →

More from ikawrakow

All by ikawrakow →

More Llama models

All Llama →

Other medium models

All medium models →