MI

Mixtral-8x22B-Instruct-v0.1

MistralGGUF by LiteLLMsUpdated 2024-05-29
Download on Hugging Face
Parameters141B
Context64K
Layers56
Architecturellama
Smallest quant52.1 GB
Quick answer

Mixtral-8x22B-Instruct-v0.1 needs about 88.0 GB of memory at Q4_K_M (a 85.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 14 quants, from 52.1 GB (Q2_K) to 149.4 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 7.5 GB.

Mixtral-8x22B-Instruct-v0.1 is a 141B-parameter Mistral model in GGUF format, quantized by LiteLLMs. Its trained context length is 64K tokens. The repository has 14 quant files, from 52.1 GB (Q2_K) to 149.4 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K52.1 GB (2 parts) 54.5 GB
Too big Get file →
Q3_K_S61.5 GB (2 parts) 63.9 GB
Too big Get file →
Q3_K_M67.8 GB (3 parts) 70.2 GB
Too big Get file →
Q4_079.5 GB (3 parts) 81.9 GB
Too big Get file →
Q4_K_S80.5 GB (3 parts) 82.9 GB
Too big Get file →
Q4_K_M85.6 GB (3 parts) 88.0 GB
Too big Get file →
Q4_188.3 GB (3 parts) 90.6 GB
Too big Get file →
Q3_K_L90.8 GB (4 parts) 93.1 GB
Too big Get file →
Q5_097.0 GB (4 parts) 99.4 GB
Too big Get file →
Q5_K_S97.0 GB (4 parts) 99.4 GB
Too big Get file →
Q5_K_M100.0 GB (4 parts) 102.4 GB
Too big Get file →
Q5_1105.7 GB (4 parts) 108.1 GB
Too big Get file →
Q6_K115.5 GB (4 parts) 117.9 GB
Too big Get file →
Q8_0149.4 GB (5 parts) 151.8 GB
Too big Get file →

What GPU do you need for Mixtral-8x22B-Instruct-v0.1?

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: no quant fits at an 8K context.
  • 48 GB VRAM: no quant fits at an 8K context.

Which Mixtral-8x22B-Instruct-v0.1 quant should you choose?

Q4_K_M is the usual starting point for Mixtral-8x22B-Instruct-v0.1: it balances file size and quality well. Here it is a 85.6 GB file.

The Q8_0 file is 149.4 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 Q2_K at 52.1 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
Parameters141B (approximate, mixture of experts)
Layers56
Trained context65,536 tokens
KV cache (fp16)224 KB per token

Source

Quantized byLiteLLMs
FamilyMistral
Licenseapache-2.0
Downloads (HF)84
Last updated2024-05-29

Frequently asked questions about Mixtral-8x22B-Instruct-v0.1

How much VRAM does Mixtral-8x22B-Instruct-v0.1 need?

About 88.0 GB at Q4_K_M with an 8K context window. That is the 85.6 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 Mixtral-8x22B-Instruct-v0.1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 54.5 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Mixtral-8x22B-Instruct-v0.1 GGUF?

It depends on the quant: Q2_K is 52.1 GB, Q3_K_S is 61.5 GB, Q3_K_M is 67.8 GB, Q4_0 is 79.5 GB, Q4_K_S is 80.5 GB, Q4_K_M is 85.6 GB and 8 more.

What context length does Mixtral-8x22B-Instruct-v0.1 support?

The GGUF header lists a trained context length of 65,536 tokens (64K). The KV cache grows with the context you use: about 1.9 GB at 8K and 7.5 GB at 32K.

Where can I download Mixtral-8x22B-Instruct-v0.1 GGUF?

The files are on Hugging Face in the LiteLLMs/Mixtral-8x22B-Instruct-v0.1-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 Mixtral-8x22B-Instruct-v0.1 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.

Similar models

Browse all →

More from LiteLLMs

All by LiteLLMs →

More Mistral models

All Mistral →

Other large models

All large models →