AY

Aya Expanse 32b

AyaGGUF by bartowskiUpdated 2024-10-24
Download on Hugging Face
Parameters32.3B
Context8K
Layers40
Architecturecommand-r
Smallest quant10.3 GB
Quick answer

Aya Expanse 32b needs about 21.6 GB of memory at Q4_K_M (a 19.8 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 23 quants, from 10.3 GB (IQ2_XS) to 64.6 GB (F16).

Aya Expanse 32b – 32.3B parameter command‑r model in GGUF format

Aya Expanse 32b is a 32.3B parameter command‑r model released in GGUF format. It supports a 8192 token context window and is licensed under CC‑BY‑NC‑4.0. The model was quantized by bartowski from the CohereLabs parent model.

Good at

  • Large context handling
  • Command‑r architecture
  • High parameter count

Limits

  • No vision support
  • Requires 8k context for memory estimates
  • CC‑BY‑NC license restricts commercial use

Best for

  • Research prototyping
  • Long‑form text generation
  • Fine‑tuning on custom data

Who it’s for. Developers and researchers needing a large command‑r model for text tasks. Suitable for non‑commercial projects due to the CC‑BY‑NC license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XS10.3 GB 12.2 GB
Too big Get file →
IQ2_S10.8 GB 12.6 GB
Too big Get file →
IQ2_M11.6 GB 13.5 GB
Too big Get file →
Q2_K12.8 GB 14.7 GB
Too big Get file →
Q2_K_L13.3 GB 15.2 GB
Too big Get file →
IQ3_XS14.0 GB 15.9 GB
Too big Get file →
Q3_K_S14.7 GB 16.6 GB
Too big Get file →
IQ3_M15.2 GB 17.1 GB
Too big Get file →
Q3_K_M16.2 GB 18.1 GB
Too big Get file →
Q3_K_L17.6 GB 19.4 GB
Too big Get file →
IQ4_XS17.8 GB 19.7 GB
Too big Get file →
Q3_K_XL18.1 GB 19.9 GB
Too big Get file →
Q4_018.8 GB 20.6 GB
Too big Get file →
Q4_K_S18.8 GB 20.7 GB
Too big Get file →
Q4_K_M19.8 GB 21.6 GB
Too big Get file →
Q4_K_L20.3 GB 22.2 GB
Too big Get file →
Q5_K_S22.5 GB 24.3 GB
Too big Get file →
Q5_K_M23.1 GB 24.9 GB
Too big Get file →
Q5_K_L23.6 GB 25.4 GB
Too big Get file →
Q6_K26.5 GB 28.3 GB
Too big Get file →
Q6_K_L27.0 GB 28.9 GB
Too big Get file →
Q8_034.3 GB 36.2 GB
Too big Get file →
F1664.6 GB (2 parts) 66.4 GB
Too big Get file →

For systems with 12GB VRAM, the IQ2_XS quant (12.2GB memory at 8k context) is the smallest that fits. If you have 16GB VRAM, consider IQ2_S (12.6GB) or Q2_K (14.7GB). For 24GB VRAM, Q4_0 (20.6GB) or Q4_K_M (21.6GB) provide higher quality. Memory estimates are approximate.

What GPU do you need for Aya Expanse 32b?

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: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: the best quant that fits is IQ3_XS (14.0 GB file, about 15.9 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_L (20.3 GB file, about 22.2 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K_L (27.0 GB file, about 28.9 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (34.3 GB file, about 36.2 GB in use).

Which Aya Expanse 32b quant should you choose?

Q4_K_M is the usual starting point for Aya Expanse 32b: it balances file size and quality well. Here it is a 19.8 GB file.

The Q8_0 file is 34.3 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 IQ2_XS at 10.3 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

Architecturecommand-r
Parameters32.3B
Layers40
Trained context8,192 tokens
KV cache (fp16)160 KB per token

Source

Quantized bybartowski
FamilyAya
Downloads (HF)1,718
Last updated2024-10-24

Frequently asked questions about Aya Expanse 32b

How much VRAM does Aya Expanse 32b need?

About 21.6 GB at Q4_K_M with an 8K context window. That is the 19.8 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 Aya Expanse 32b on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 12.2 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Aya Expanse 32b GGUF?

It depends on the quant: IQ2_XS is 10.3 GB, IQ2_S is 10.8 GB, IQ2_M is 11.6 GB, Q2_K is 12.8 GB, Q2_K_L is 13.3 GB, IQ3_XS is 14.0 GB and 17 more.

What context length does Aya Expanse 32b support?

The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 1.3 GB at 8K and 1.3 GB at 8K.

Where can I download Aya Expanse 32b GGUF?

The files are on Hugging Face in the bartowski/aya-expanse-32b-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 Aya Expanse 32b use?

Hugging Face lists the license as cc-by-nc-4.0. 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 Aya Expanse 32b on 8GB VRAM?

No, the smallest quant requires at least 12.2GB memory at 8k context.

Does Aya Expanse 32b support vision?

No vision support is available for this model.

Other GGUFs of aya-expanse-32b

Similar models

Browse all →

More from bartowski

All by bartowski →

More Aya models

All Aya →

Other medium models

All medium models →