SH

shqiponja 15b v1

ShqiponjaGGUF by Artefact2Updated 2024-02-07
Download on Hugging Face
Parameters15.6B
Context32K
Layers39
Architecturellama
Smallest quant4.2 GB
Quick answer

shqiponja 15b v1 needs about 11.2 GB of memory at Q4_K_M (a 9.4 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 12 quants, from 4.2 GB (IQ2_XXS) to 12.8 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 5.2 GB.

shqiponja 15b v1 is a 15.6B-parameter Shqiponja model in GGUF format, quantized by Artefact2. Its trained context length is 32K tokens. The repository has 12 quant files, from 4.2 GB (IQ2_XXS) to 12.8 GB (Q6_K).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XXS4.2 GB 6.0 GB
Fits Get file →
IQ2_XS4.6 GB 6.5 GB
Fits Get file →
Q2_K_S5.3 GB 7.2 GB
Fits Get file →
Q2_K5.8 GB 7.6 GB
Tight Get file →
IQ3_XXS6.1 GB 7.9 GB
Tight Get file →
Q3_K_M7.5 GB 9.3 GB
Too big Get file →
Q3_K_L8.2 GB 10.0 GB
Too big Get file →
Q4_K_S8.9 GB 10.7 GB
Too big Get file →
Q4_K_M9.4 GB 11.2 GB
Too big Get file →
Q5_K_S10.8 GB 12.6 GB
Too big Get file →
Q5_K_M11.1 GB 12.9 GB
Too big Get file →
Q6_K12.8 GB 14.6 GB
Too big Get file →

What GPU do you need for shqiponja 15b v1?

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: the best quant that fits is IQ2_XXS (4.2 GB file, about 6.0 GB in use).
  • 8 GB VRAM: the best quant that fits is IQ3_XXS (6.1 GB file, about 7.9 GB in use).
  • 12 GB VRAM: the best quant that fits is Q4_K_M (9.4 GB file, about 11.2 GB in use).
  • 16 GB VRAM: the best quant that fits is Q6_K (12.8 GB file, about 14.6 GB in use).

Which shqiponja 15b v1 quant should you choose?

Q4_K_M is the usual starting point for shqiponja 15b v1: it balances file size and quality well. Here it is a 9.4 GB file.

If memory is tight, the smallest file is IQ2_XXS at 4.2 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters15.6B (approximate, mixture of experts)
Layers39
Trained context32,768 tokens
KV cache (fp16)156 KB per token

Source

Quantized byArtefact2
FamilyShqiponja
Licensegpl-3.0
Downloads (HF)37
Last updated2024-02-07

Frequently asked questions about shqiponja 15b v1

How much VRAM does shqiponja 15b v1 need?

About 11.2 GB at Q4_K_M with an 8K context window. That is the 9.4 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 shqiponja 15b v1 on an 8 GB GPU?

Yes, with IQ3_XXS: the 6.1 GB file needs about 7.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of shqiponja 15b v1 GGUF?

It depends on the quant: IQ2_XXS is 4.2 GB, IQ2_XS is 4.6 GB, Q2_K_S is 5.3 GB, Q2_K is 5.8 GB, IQ3_XXS is 6.1 GB, Q3_K_M is 7.5 GB and 6 more.

What context length does shqiponja 15b v1 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.3 GB at 8K and 5.2 GB at 32K.

Where can I download shqiponja 15b v1 GGUF?

The files are on Hugging Face in the Artefact2/shqiponja-15b-v1-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 shqiponja 15b v1 use?

Hugging Face lists the license as gpl-3.0. 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 Artefact2

All by Artefact2 →

More Shqiponja models

All Shqiponja →

Other medium models

All medium models →