GE

firefly-gemma-7b

GemmaGGUF by LewdiculousUpdated 2024-03-20
Download on Hugging Face
Parameters8.54B
Context8K
Layers28
Architecturegemma
Smallest quant2.6 GB
Quick answer

firefly-gemma-7b needs about 8.4 GB of memory at IQ3_M (a 4.1 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 16 quants, from 2.6 GB (IQ2_XXS) to 17.1 GB (F16).

firefly-gemma-7b is a 8.54B-parameter Gemma model in GGUF format, quantized by Lewdiculous. Its trained context length is 8K tokens. The repository has 16 quant files, from 2.6 GB (IQ2_XXS) to 17.1 GB (F16).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XXS2.6 GB 6.8 GB
Fits Get file →
IQ2_XS2.8 GB 7.1 GB
Fits Get file →
IQ2_S2.9 GB 7.2 GB
Fits Get file →
IQ2_M3.1 GB 7.4 GB
Tight Get file →
IQ3_XXS3.5 GB 7.7 GB
Tight Get file →
IQ3_XS3.8 GB 8.1 GB
Too big Get file →
IQ3_S4.0 GB 8.2 GB
Too big Get file →
IQ3_M4.1 GB 8.4 GB
Too big Get file →
Q3_K_M4.4 GB 8.6 GB
Too big Get file →
IQ4_XS4.8 GB 9.0 GB
Too big Get file →
Q4_K_S5.0 GB 9.3 GB
Too big Get file →
Q5_K_S6.0 GB 10.2 GB
Too big Get file →
Q5_K_M6.1 GB 10.4 GB
Too big Get file →
Q6_K7.0 GB 11.3 GB
Too big Get file →
Q8_09.1 GB 13.3 GB
Too big Get file →
F1617.1 GB 21.3 GB
Too big Get file →

What GPU do you need for firefly-gemma-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 IQ3_XXS (3.5 GB file, about 7.7 GB in use).
  • 12 GB VRAM: the best quant that fits is Q6_K (7.0 GB file, about 11.3 GB in use).
  • 16 GB VRAM: the best quant that fits is Q8_0 (9.1 GB file, about 13.3 GB in use).
  • 24 GB VRAM: the best quant that fits is F16 (17.1 GB file, about 21.3 GB in use).

Which firefly-gemma-7b quant should you choose?

IQ3_M (4.1 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

If memory is tight, the smallest file is IQ2_XXS at 2.6 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

Architecturegemma
Parameters8.54B
Layers28
Trained context8,192 tokens
KV cache (fp16)448 KB per token

Source

Quantized byLewdiculous
FamilyGemma
Licenseapache-2.0
Downloads (HF)130
Last updated2024-03-20

Frequently asked questions about firefly-gemma-7b

How much VRAM does firefly-gemma-7b need?

About 8.4 GB at IQ3_M 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 firefly-gemma-7b on an 8 GB GPU?

Yes, with IQ3_XXS: the 3.5 GB file needs about 7.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of firefly-gemma-7b GGUF?

It depends on the quant: IQ2_XXS is 2.6 GB, IQ2_XS is 2.8 GB, IQ2_S is 2.9 GB, IQ2_M is 3.1 GB, IQ3_XXS is 3.5 GB, IQ3_XS is 3.8 GB and 10 more.

What context length does firefly-gemma-7b support?

The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 3.8 GB at 8K and 3.8 GB at 8K.

Where can I download firefly-gemma-7b GGUF?

The files are on Hugging Face in the Lewdiculous/firefly-gemma-7b-GGUF-IQ-Imatrix repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does firefly-gemma-7b 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 Lewdiculous

All by Lewdiculous →

More Gemma models

All Gemma →

Other small models

All small models →