QW

Qwen2.5 32B

QwenGGUF by bartowskiUpdated 2024-10-23
Download on Hugging Face
Parameters32.8B
Context128K
Layers64
Architectureqwen2
Smallest quant9.0 GB
Quick answer

Qwen2.5 32B needs about 22.5 GB of memory at Q4_K_M (a 19.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 27 quants, from 9.0 GB (IQ2_XXS) to 65.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.

EVA-Qwen2.5-32B v0.0 – a 32.8B parameter role‑play and story‑writing model

EVA-Qwen2.5-32B v0.0 is a 32.8B parameter model fine‑tuned from Qwen2.5‑32B for role‑play and story writing. It supports a 131072 token context window and uses the ChatML prompt format. The model is licensed under apache‑2.0.

Good at

  • Role‑play and story writing
  • ChatML prompt handling
  • Large context support

Limits

  • No vision support
  • Quantized KV cache not recommended
  • Requires 8k context for memory estimates

Best for

  • Creative writing assistants
  • Interactive storytelling bots
  • ChatML‑based applications

Who it’s for. Developers building role‑play or story‑writing applications who need a large‑context model. Users who prefer open‑source models under an apache‑2.0 license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XXS9.0 GB 11.7 GB
Too big Get file →
IQ2_XS10.0 GB 12.6 GB
Too big Get file →
IQ2_S10.4 GB 13.0 GB
Too big Get file →
IQ2_M11.3 GB 13.9 GB
Too big Get file →
Q2_K12.3 GB 15.0 GB
Too big Get file →
Q2_K_L13.1 GB 15.7 GB
Too big Get file →
IQ3_XS13.7 GB 16.4 GB
Too big Get file →
Q3_K_S14.4 GB 17.0 GB
Too big Get file →
IQ3_M14.8 GB 17.5 GB
Too big Get file →
Q3_K_M15.9 GB 18.6 GB
Too big Get file →
Q3_K_L17.2 GB 19.9 GB
Too big Get file →
IQ4_XS17.7 GB 20.3 GB
Too big Get file →
Q3_K_XL17.9 GB 20.6 GB
Too big Get file →
Q4_018.6 GB 21.3 GB
Too big Get file →
Q4_018.6 GB 21.3 GB
Too big Get file →
Q4_018.6 GB 21.3 GB
Too big Get file →
Q4_018.7 GB 21.4 GB
Too big Get file →
Q4_K_S18.8 GB 21.4 GB
Too big Get file →
Q4_K_M19.9 GB 22.5 GB
Too big Get file →
Q4_K_L20.4 GB 23.1 GB
Too big Get file →
Q5_K_S22.6 GB 25.3 GB
Too big Get file →
Q5_K_M23.3 GB 25.9 GB
Too big Get file →
Q5_K_L23.7 GB 26.4 GB
Too big Get file →
Q6_K26.9 GB 29.5 GB
Too big Get file →
Q6_K_L27.3 GB 29.9 GB
Too big Get file →
Q8_034.8 GB 37.5 GB
Too big Get file →
F1665.5 GB (2 parts) 68.2 GB
Too big Get file →

For 8‑GB VRAM use IQ2_XXS (9.03 GB file, 11.7 GB memory). For 16‑GB VRAM use Q4_K_M (19.85 GB file, 22.5 GB memory). For 24‑GB VRAM use Q5_K_L (23.74 GB file, 26.4 GB memory). Estimates are approximate.

What GPU do you need for Qwen2.5 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: the best quant that fits is IQ2_XXS (9.0 GB file, about 11.7 GB in use).
  • 16 GB VRAM: the best quant that fits is Q2_K_L (13.1 GB file, about 15.7 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_K_L (20.4 GB file, about 23.1 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K_L (27.3 GB file, about 29.9 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (34.8 GB file, about 37.5 GB in use).

Which Qwen2.5 32B quant should you choose?

Q4_K_M is the usual starting point for Qwen2.5 32B: it balances file size and quality well. Here it is a 19.9 GB file.

The Q8_0 file is 34.8 GB, about 1.8 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_XXS at 9.0 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

Architectureqwen2
Parameters32.8B
Layers64
Trained context131,072 tokens
KV cache (fp16)256 KB per token

Source

Quantized bybartowski
FamilyQwen
Licenseapache-2.0
Downloads (HF)1,105
Last updated2024-10-23

Frequently asked questions about Qwen2.5 32B

How much VRAM does Qwen2.5 32B need?

About 22.5 GB at Q4_K_M with an 8K context window. That is the 19.9 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 Qwen2.5 32B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 11.7 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Qwen2.5 32B GGUF?

It depends on the quant: IQ2_XXS is 9.0 GB, IQ2_XS is 10.0 GB, IQ2_S is 10.4 GB, IQ2_M is 11.3 GB, Q2_K is 12.3 GB, Q2_K_L is 13.1 GB and 21 more.

What context length does Qwen2.5 32B support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.6 GB at 32K.

Where can I download Qwen2.5 32B GGUF?

The files are on Hugging Face in the bartowski/EVA-Qwen2.5-32B-v0.0-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 Qwen2.5 32B 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.

Can I run Qwen2.5‑32B on 8 GB VRAM?

Use the IQ2_XXS quant, which is 9.03 GB and requires about 11.7 GB of VRAM at 8k context.

Does this model support vision?

No, the model does not have vision capabilities.

What prompt format does it use?

The model uses the ChatML prompt format.

Other GGUFs of EVA-Qwen2.5-32B-v0.0

Similar models

Browse all →

More from bartowski

All by bartowski →

More Qwen models

All Qwen →

Other medium models

All medium models →