MO

MoNeuTrix MoE 4x7B i1

MoneutrixGGUF by mradermacherUpdated 2024-05-06
Download on Hugging Face
Parameters24.2B
Context32K
Layers32
Architecturellama
Smallest quant5.2 GB
Quick answer

MoNeuTrix MoE 4x7B i1 needs about 16.3 GB of memory at Q4_K_M (a 14.8 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 22 quants, from 5.2 GB (IQ1_S) to 20.0 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.

MoNeuTrix MoE 4x7B i1 – 24.2B parameter Mixture of Experts model

MoNeuTrix MoE 4x7B i1 is a 24.2B parameter Mixture of Experts model based on the llama architecture. It was trained with a context length of 32768 tokens and is released under the apache-2.0 license. The model is available in multiple GGUF quantizations.

Good at

  • Large context handling
  • Mixture of Experts efficiency
  • High parameter count

Limits

  • No vision support
  • Requires large VRAM for higher quants
  • No active parameters listed

Best for

  • Long‑form text generation
  • Complex reasoning tasks
  • Research prototyping

Who it’s for. Developers and researchers needing a large, context‑rich language model with Mixture of Experts capabilities. Suitable for environments that can accommodate the memory requirements of the chosen quantization.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S5.2 GB 6.8 GB
Fits Get file →
IQ1_M5.7 GB 7.3 GB
Tight Get file →
IQ2_XXS6.6 GB 8.1 GB
Too big Get file →
IQ2_XS7.3 GB 8.9 GB
Too big Get file →
IQ2_S7.5 GB 9.0 GB
Too big Get file →
IQ2_M8.2 GB 9.7 GB
Too big Get file →
Q2_K9.0 GB 10.6 GB
Too big Get file →
IQ3_XXS9.5 GB 11.0 GB
Too big Get file →
IQ3_XS10.0 GB 11.6 GB
Too big Get file →
Q3_K_S10.6 GB 12.2 GB
Too big Get file →
IQ3_S10.6 GB 12.2 GB
Too big Get file →
IQ3_M10.8 GB 12.4 GB
Too big Get file →
Q3_K_M11.7 GB 13.3 GB
Too big Get file →
Q3_K_L12.7 GB 14.3 GB
Too big Get file →
IQ4_XS13.0 GB 14.6 GB
Too big Get file →
IQ4_NL13.8 GB 15.4 GB
Too big Get file →
Q4_013.8 GB 15.4 GB
Too big Get file →
Q4_K_S13.9 GB 15.5 GB
Too big Get file →
Q4_K_M14.8 GB 16.3 GB
Too big Get file →
Q5_K_S16.8 GB 18.4 GB
Too big Get file →
Q5_K_M17.3 GB 18.9 GB
Too big Get file →
Q6_K20.0 GB 21.5 GB
Too big Get file →

For 8 GB VRAM, use IQ1_S (file size 5.18 GB, memory 6.8 GB) or IQ1_M (5.7 GB, 7.3 GB). For 12 GB VRAM, IQ2_S (7.47 GB, 9 GB) or IQ2_M (8.17 GB, 9.7 GB) are suitable. Higher quants such as Q4_K_M (14.76 GB, 16.3 GB) require more than 16 GB VRAM.

What GPU do you need for MoNeuTrix MoE 4x7B i1?

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 IQ1_M (5.7 GB file, about 7.3 GB in use).
  • 12 GB VRAM: the best quant that fits is IQ3_XS (10.0 GB file, about 11.6 GB in use).
  • 16 GB VRAM: the best quant that fits is Q4_K_S (13.9 GB file, about 15.5 GB in use).
  • 24 GB VRAM: the best quant that fits is Q6_K (20.0 GB file, about 21.5 GB in use).

Which MoNeuTrix MoE 4x7B i1 quant should you choose?

Q4_K_M is the usual starting point for MoNeuTrix MoE 4x7B i1: it balances file size and quality well. Here it is a 14.8 GB file.

If memory is tight, the smallest file is IQ1_S at 5.2 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

Architecturellama
Parameters24.2B (approximate, mixture of experts)
Layers32
Trained context32,768 tokens
KV cache (fp16)128 KB per token

Source

Quantized bymradermacher
FamilyMoneutrix
Licenseapache-2.0
Downloads (HF)2,651
Last updated2024-05-06

Frequently asked questions about MoNeuTrix MoE 4x7B i1

How much VRAM does MoNeuTrix MoE 4x7B i1 need?

About 16.3 GB at Q4_K_M with an 8K context window. That is the 14.8 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 MoNeuTrix MoE 4x7B i1 on an 8 GB GPU?

Yes, with IQ1_M: the 5.7 GB file needs about 7.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of MoNeuTrix MoE 4x7B i1 GGUF?

It depends on the quant: IQ1_S is 5.2 GB, IQ1_M is 5.7 GB, IQ2_XXS is 6.6 GB, IQ2_XS is 7.3 GB, IQ2_S is 7.5 GB, IQ2_M is 8.2 GB and 16 more.

What context length does MoNeuTrix MoE 4x7B i1 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.

Where can I download MoNeuTrix MoE 4x7B i1 GGUF?

The files are on Hugging Face in the mradermacher/MoNeuTrix-MoE-4x7B-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 MoNeuTrix MoE 4x7B i1 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 MoNeuTrix MoE 4x7B i1 on 8 GB VRAM?

Yes, using the IQ1_S or IQ1_M quantizations which use 6.8 GB or 7.3 GB of VRAM at 8k context.

What license does this model use?

The model is released under the apache-2.0 license.

Does this model support vision tasks?

No, the model does not have vision capabilities.

Other GGUFs of MoNeuTrix-MoE-4x7B

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Moneutrix models

All Moneutrix →

Other medium models

All medium models →