QW

Qwen1.5 110B

QwenGGUF by mradermacherUpdated 2024-09-18
Download on Hugging Face
Parameters111B
Context32K
Layers80
Architectureqwen2
Smallest quant23.6 GB
Quick answer

Qwen1.5 110B needs about 40.6 GB of memory at IQ2_M (a 37.4 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 12 quants, from 23.6 GB (IQ1_S) to 49.7 GB (IQ3_M). Longer context needs more: at 32K tokens the KV cache alone adds about 10.7 GB.

Qwen1.5 110B – 111B parameter decoder-only language model

Qwen1.5 110B is a 111B parameter transformer decoder model. It supports a 32768 token context length and is part of the Qwen2 family. The model is available in multiple GGUF quantizations.

Good at

  • Large context generation
  • High‑parameter language modeling
  • Multilingual text support

Limits

  • Requires large VRAM for full precision
  • No vision capabilities
  • License type is other

Best for

  • Research on large language models
  • Fine‑tuning for specialized domains
  • Benchmarking quantization techniques

Who it’s for. Researchers and developers needing a very large language model for experimentation. The model is suitable for those who can allocate sufficient GPU memory and are comfortable with GGUF quantizations.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S23.6 GB 26.8 GB
Too big Get file →
IQ1_M25.9 GB 29.1 GB
Too big Get file →
IQ2_XXS29.8 GB 33.0 GB
Too big Get file →
IQ2_XS33.0 GB 36.2 GB
Too big Get file →
IQ2_S34.3 GB 37.5 GB
Too big Get file →
IQ2_M37.4 GB 40.6 GB
Too big Get file →
Q2_K41.2 GB 44.4 GB
Too big Get file →
IQ3_XXS43.1 GB 46.3 GB
Too big Get file →
IQ3_XS45.9 GB 49.1 GB
Too big Get file →
IQ3_S48.5 GB 51.6 GB
Too big Get file →
Q3_K_S48.5 GB 51.6 GB
Too big Get file →
IQ3_M49.7 GB 52.9 GB
Too big Get file →

For 8‑GB VRAM, use IQ1_S (23.63 GB file, 26.8 GB memory at 8K context). For 12‑GB VRAM, IQ1_M (25.94 GB file, 29.1 GB memory). For 16‑GB VRAM, IQ2_XXS (29.8 GB file, 33 GB memory).

What GPU do you need for Qwen1.5 110B?

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: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is IQ1_M (25.9 GB file, about 29.1 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_XXS (43.1 GB file, about 46.3 GB in use).

Which Qwen1.5 110B quant should you choose?

IQ2_M (37.4 GB) is the middle option in this repository. 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.

If memory is tight, the smallest file is IQ1_S at 23.6 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

Architectureqwen2
Parameters111B
Layers80
Trained context32,768 tokens
KV cache (fp16)320 KB per token

Source

Quantized bymradermacher
FamilyQwen
Licenseother
Downloads (HF)697
Last updated2024-09-18

Frequently asked questions about Qwen1.5 110B

How much VRAM does Qwen1.5 110B need?

About 40.6 GB at IQ2_M with an 8K context window. That is the 37.4 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 Qwen1.5 110B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 26.8 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Qwen1.5 110B GGUF?

It depends on the quant: IQ1_S is 23.6 GB, IQ1_M is 25.9 GB, IQ2_XXS is 29.8 GB, IQ2_XS is 33.0 GB, IQ2_S is 34.3 GB, IQ2_M is 37.4 GB and 6 more.

What context length does Qwen1.5 110B support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 2.7 GB at 8K and 10.7 GB at 32K.

Where can I download Qwen1.5 110B GGUF?

The files are on Hugging Face in the mradermacher/Qwen1.5-110B-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 Qwen1.5 110B use?

Hugging Face lists the license as other. 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 Qwen1.5 110B on 8 GB VRAM?

The smallest quant, IQ1_S, requires 26.8 GB memory at 8K context, so 8 GB VRAM is insufficient.

Is there a quant that fits 12 GB VRAM?

IQ1_M uses 29.1 GB memory at 8K context, which exceeds 12 GB VRAM; a smaller quant would be needed.

Does the model support vision tasks?

No, the model does not include vision capabilities.

Other GGUFs of Qwen1.5-110B

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Qwen models

All Qwen →

Other large models

All large models →