SA

saiga2 7b

SaigaGGUF by IlyaGusevUpdated 2023-09-17
Download on Hugging Face
Parameters6.74B
Context4K
Layers32
Architecturellama
Smallest quant2.8 GB
Quick answer

saiga2 7b needs about 8.9 GB of memory at Q4_K (a 4.1 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 5 quants, from 2.8 GB (Q2_K) to 7.2 GB (Q8_0).

saiga2 7b is a 6.74B-parameter Saiga model in GGUF format, quantized by IlyaGusev. Its trained context length is 4K tokens. The repository has 5 quant files, from 2.8 GB (Q2_K) to 7.2 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K2.8 GB 5.5 GB
Fits Get file →
Q3_K3.3 GB 5.9 GB
Fits Get file →
Q4_K4.1 GB 6.7 GB
Fits Get file →
Q5_K4.8 GB 7.4 GB
Tight Get file →
Q8_07.2 GB 9.8 GB
Too big Get file →

What GPU do you need for saiga2 7b?

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 Q2_K (2.8 GB file, about 7.6 GB in use).
  • 12 GB VRAM: the best quant that fits is Q8_0 (7.2 GB file, about 12.0 GB in use).

Which saiga2 7b quant should you choose?

Q4_K (4.1 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is Q2_K at 2.8 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
Parameters6.74B
Layers32
Trained context4,096 tokens
KV cache (fp16)512 KB per token

Source

Quantized byIlyaGusev
FamilySaiga
Licensellama2
Downloads (HF)526
Last updated2023-09-17

Frequently asked questions about saiga2 7b

How much VRAM does saiga2 7b need?

About 8.9 GB at Q4_K with an 8K context window. That is the 4.1 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 saiga2 7b on an 8 GB GPU?

Yes, with Q2_K: the 2.8 GB file needs about 7.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of saiga2 7b GGUF?

It depends on the quant: Q2_K is 2.8 GB, Q3_K is 3.3 GB, Q4_K is 4.1 GB, Q5_K is 4.8 GB, Q8_0 is 7.2 GB.

What context length does saiga2 7b 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 saiga2 7b GGUF?

The files are on Hugging Face in the IlyaGusev/saiga2_7b_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 saiga2 7b 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.

Similar models

Browse all →

More from IlyaGusev

All by IlyaGusev →

More Saiga models

All Saiga →

Other small models

All small models →