LL

Llama 3.2 3B Instruct

LlamaGGUF by velyanUpdated 2024-11-02
Download on Hugging Face
Parameters1.24B
Context128K
Layers28
Architecturellama
Smallest quant0.7 GB
Quick answer

Llama 3.2 3B Instruct needs about 3.5 GB of memory at Q4_K_M (a 2.0 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 7 quants, from 0.7 GB (IQ3_M) to 2.3 GB (Q5_K_S). Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.

Llama 3.2 3B Instruct is a 1.24B-parameter Llama model in GGUF format, quantized by velyan. Its trained context length is 128K tokens. The repository has 7 quant files, from 0.7 GB (IQ3_M) to 2.3 GB (Q5_K_S).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ3_M0.7 GB 2.1 GB
Fits Get file →
Q5_K_M0.9 GB 2.4 GB
Fits Get file →
IQ3_XXS1.3 GB 2.8 GB
Fits Get file →
IQ3_M1.6 GB 3.0 GB
Fits Get file →
Q4_K_M2.0 GB 3.5 GB
Fits Get file →
—2.0 GB 3.5 GB
Fits Get file →
Q5_K_S2.3 GB 3.7 GB
Fits Get file →

What GPU do you need for Llama 3.2 3B Instruct?

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: the best quant that fits is Q5_K_S (2.3 GB file, about 3.7 GB in use).

Which Llama 3.2 3B Instruct quant should you choose?

Q4_K_M is the usual starting point for Llama 3.2 3B Instruct: it balances file size and quality well. Here it is a 2.0 GB file.

If memory is tight, the smallest file is IQ3_M at 0.7 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters1.24B
Layers28
Trained context131,072 tokens
KV cache (fp16)112 KB per token

Source

Quantized byvelyan
FamilyLlama
Licensellama3.2
Downloads (HF)135
Last updated2024-11-02

Frequently asked questions about Llama 3.2 3B Instruct

How much VRAM does Llama 3.2 3B Instruct need?

About 3.5 GB at Q4_K_M with an 8K context window. That is the 2.0 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 Llama 3.2 3B Instruct on an 8 GB GPU?

Yes, with Q5_K_S: the 2.3 GB file needs about 3.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Llama 3.2 3B Instruct GGUF?

It depends on the quant: IQ3_M is 0.7 GB, Q5_K_M is 0.9 GB, IQ3_XXS is 1.3 GB, IQ3_M is 1.6 GB, Q4_K_M is 2.0 GB, file is 2.0 GB and 1 more.

What context length does Llama 3.2 3B Instruct support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.8 GB at 32K.

Where can I download Llama 3.2 3B Instruct GGUF?

The files are on Hugging Face in the velyan/Llama-3.2-3B-Instruct-Q4_K_M-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 3.2 3B Instruct use?

Hugging Face lists the license as llama3.2. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Other GGUFs of Llama-3.2-3B-Instruct

Similar models

Browse all →

More from velyan

All by velyan →

More Llama models

All Llama →

Other tiny models

All tiny models →