GE

gemma-2-9b

GemmaGGUF by mradermacherUpdated 2024-08-02
Download on Hugging Face
Parameters9.24B
Context8K
Layers42
Architecturegemma2
Smallest quant2.4 GB
Quick answer

gemma-2-9b needs about 9.1 GB of memory at Q4_K_M (a 5.8 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 21 quants, from 2.4 GB (IQ1_S) to 7.6 GB (Q6_K).

Gemma‑2‑9b in GGUF format

Gemma‑2‑9b is a 9.24‑billion‑parameter model from the Gemma family. It supports a 8192‑token context window and is available in a variety of quantized GGUF files. The model is released under the gemma license.

Good at

  • Fast inference on modest GPUs
  • Supports 8192‑token context
  • Available in many quantizations

Limits

  • No vision support
  • License is gemma
  • No mixture of experts

Best for

  • Chatbots with 8k context
  • Text generation with low VRAM
  • Fine‑tuning on small datasets

Who it’s for. Developers and researchers who need a large language model that can run on consumer GPUs. The model is suitable for applications that require a long context window and moderate memory usage.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S2.4 GB 5.7 GB
Fits Get file →
IQ1_M2.5 GB 5.9 GB
Fits Get file →
IQ2_XXS2.8 GB 6.1 GB
Fits Get file →
IQ2_XS3.1 GB 6.4 GB
Fits Get file →
IQ2_S3.2 GB 6.5 GB
Fits Get file →
IQ2_M3.4 GB 6.8 GB
Fits Get file →
IQ3_XXS3.8 GB 7.1 GB
Fits Get file →
Q2_K3.8 GB 7.1 GB
Fits Get file →
IQ3_XS4.1 GB 7.5 GB
Tight Get file →
IQ3_S4.3 GB 7.7 GB
Tight Get file →
Q3_K_S4.3 GB 7.7 GB
Tight Get file →
IQ3_M4.5 GB 7.8 GB
Tight Get file →
Q3_K_M4.8 GB 8.1 GB
Too big Get file →
Q3_K_L5.1 GB 8.5 GB
Too big Get file →
IQ4_XS5.2 GB 8.5 GB
Too big Get file →
Q4_05.5 GB 8.8 GB
Too big Get file →
Q4_K_S5.5 GB 8.8 GB
Too big Get file →
Q4_K_M5.8 GB 9.1 GB
Too big Get file →
Q5_K_S6.5 GB 9.8 GB
Too big Get file →
Q5_K_M6.6 GB 10.0 GB
Too big Get file →
Q6_K7.6 GB 10.9 GB
Too big Get file →

For 8‑GB VRAM, use IQ1_S (2.38 GB file, 5.7 GB memory) or IQ1_M (2.55 GB file, 5.9 GB memory). If 12 GB VRAM is available, IQ2_S (3.21 GB file, 6.5 GB memory) or Q2_K (3.81 GB file, 7.1 GB memory) are good options. For 16 GB VRAM, Q4_K_M (5.76 GB file, 9.1 GB memory) or Q5_K_M (6.65 GB file, 10 GB memory) provide higher quality.

What GPU do you need for gemma-2-9b?

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 IQ1_M (2.5 GB file, about 5.9 GB in use).
  • 8 GB VRAM: the best quant that fits is IQ3_M (4.5 GB file, about 7.8 GB in use).
  • 12 GB VRAM: the best quant that fits is Q6_K (7.6 GB file, about 10.9 GB in use).

Which gemma-2-9b quant should you choose?

Q4_K_M is the usual starting point for gemma-2-9b: it balances file size and quality well. Here it is a 5.8 GB file.

If memory is tight, the smallest file is IQ1_S at 2.4 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

Architecturegemma2
Parameters9.24B
Layers42
Trained context8,192 tokens
KV cache (fp16)336 KB per token

Source

Quantized bymradermacher
FamilyGemma
Licensegemma
Downloads (HF)913
Last updated2024-08-02

Frequently asked questions about gemma-2-9b

How much VRAM does gemma-2-9b need?

About 9.1 GB at Q4_K_M with an 8K context window. That is the 5.8 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 gemma-2-9b on an 8 GB GPU?

Yes, with IQ3_M: the 4.5 GB file needs about 7.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of gemma-2-9b GGUF?

It depends on the quant: IQ1_S is 2.4 GB, IQ1_M is 2.5 GB, IQ2_XXS is 2.8 GB, IQ2_XS is 3.1 GB, IQ2_S is 3.2 GB, IQ2_M is 3.4 GB and 15 more.

What context length does gemma-2-9b support?

The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 2.8 GB at 8K and 2.8 GB at 8K.

Where can I download gemma-2-9b GGUF?

The files are on Hugging Face in the mradermacher/gemma-2-9b-i1-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 gemma-2-9b use?

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

Can I run this model on 8 GB VRAM?

Yes, use IQ1_S or IQ1_M quantizations, which require about 5.7–5.9 GB memory at 8k context.

Does this model support vision tasks?

No, the model has no vision capabilities.

What license applies to this model?

The model is released under the gemma license.

Other GGUFs of gemma-2-9b

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Gemma models

All Gemma →

Other small models

All small models →