LL

Pruned50 Llama 3.2 3B

LlamaGGUF by mikepwcscalingUpdated 2025-02-06
Download on Hugging Face
Parameters2.16B
Context128K
Layers28
Architecturellama
Smallest quant1.3 GB
Quick answer

Pruned50 Llama 3.2 3B needs about 2.8 GB of memory at Q4_K_M (a 1.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 3 quants, from 1.3 GB (Q4_0) to 2.3 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.

Pruned50 Llama 3.2 3B is a 2.16B-parameter Llama model in GGUF format, quantized by mikepwcscaling. Its trained context length is 128K tokens. The repository has 3 quant files, from 1.3 GB (Q4_0) to 2.3 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_01.3 GB 2.8 GB
Fits Get file →
Q4_K_M1.4 GB 2.8 GB
Fits Get file →
Q8_02.3 GB 3.7 GB
Fits Get file →

What GPU do you need for Pruned50 Llama 3.2 3B?

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 Q8_0 (2.3 GB file, about 3.7 GB in use).

Which Pruned50 Llama 3.2 3B quant should you choose?

Q4_K_M is the usual starting point for Pruned50 Llama 3.2 3B: it balances file size and quality well. Here it is a 1.4 GB file.

The Q8_0 file is 2.3 GB, about 1.7 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 1.3 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

Architecturellama
Parameters2.16B
Layers28
Trained context131,072 tokens
KV cache (fp16)112 KB per token

Source

Quantized bymikepwcscaling
FamilyLlama
Licenseapache-2.0
Downloads (HF)10
Last updated2025-02-06

Frequently asked questions about Pruned50 Llama 3.2 3B

How much VRAM does Pruned50 Llama 3.2 3B need?

About 2.8 GB at Q4_K_M with an 8K context window. That is the 1.4 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 Pruned50 Llama 3.2 3B on an 8 GB GPU?

Yes, with Q8_0: 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 Pruned50 Llama 3.2 3B GGUF?

It depends on the quant: Q4_0 is 1.3 GB, Q4_K_M is 1.4 GB, Q8_0 is 2.3 GB.

What context length does Pruned50 Llama 3.2 3B 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 Pruned50 Llama 3.2 3B GGUF?

The files are on Hugging Face in the mikepwcscaling/pruned50-Llama-3.2-3B repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Pruned50 Llama 3.2 3B 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 mikepwcscaling

All by mikepwcscaling →

More Llama models

All Llama →

Other tiny models

All tiny models →