MI

Mixtral 8x7B Instruct v0.1 upscaled i1

MistralGGUF by mradermacherUpdated 2024-05-06
Download on Hugging Face
Parameters81.5B
Context32K
Layers56
Architecturellama
Smallest quant16.7 GB
Quick answer

Mixtral 8x7B Instruct v0.1 upscaled i1 needs about 36.0 GB of memory at IQ3_XS (a 33.6 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 17 quants, from 16.7 GB (IQ1_S) to 46.7 GB (Q4_K_S). Longer context needs more: at 32K tokens the KV cache alone adds about 7.5 GB.

Mixtral 8x7B Instruct v0.1 upscaled i1 is a 81.5B-parameter Mistral model in GGUF format, quantized by mradermacher. Its trained context length is 32K tokens. The repository has 17 quant files, from 16.7 GB (IQ1_S) to 46.7 GB (Q4_K_S).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S16.7 GB 19.1 GB
Too big Get file →
IQ1_M18.6 GB 20.9 GB
Too big Get file →
IQ2_XXS21.9 GB 24.2 GB
Too big Get file →
IQ2_S24.2 GB 26.6 GB
Too big Get file →
IQ2_XS24.3 GB 26.6 GB
Too big Get file →
IQ2_M27.0 GB 29.4 GB
Too big Get file →
Q2_K30.2 GB 32.6 GB
Too big Get file →
IQ3_XXS31.8 GB 34.2 GB
Too big Get file →
IQ3_XS33.6 GB 36.0 GB
Too big Get file →
IQ3_S35.6 GB 38.0 GB
Too big Get file →
Q3_K_S35.6 GB 38.0 GB
Too big Get file →
IQ3_M37.4 GB 39.8 GB
Too big Get file →
Q3_K_M39.3 GB 41.7 GB
Too big Get file →
Q3_K_L42.2 GB 44.5 GB
Too big Get file →
IQ4_XS43.8 GB 46.1 GB
Too big Get file →
Q4_046.3 GB 48.7 GB
Too big Get file →
Q4_K_S46.7 GB 49.0 GB
Too big Get file →

What GPU do you need for Mixtral 8x7B Instruct v0.1 upscaled 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: 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: the best quant that fits is IQ1_M (18.6 GB file, about 20.9 GB in use).
  • 32 GB VRAM: the best quant that fits is IQ2_M (27.0 GB file, about 29.4 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ4_XS (43.8 GB file, about 46.1 GB in use).

Which Mixtral 8x7B Instruct v0.1 upscaled i1 quant should you choose?

IQ3_XS (33.6 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

If memory is tight, the smallest file is IQ1_S at 16.7 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
Parameters81.5B (approximate, mixture of experts)
Layers56
Trained context32,768 tokens
KV cache (fp16)224 KB per token

Source

Quantized bymradermacher
FamilyMistral
Licenseapache-2.0
Downloads (HF)353
Last updated2024-05-06

Frequently asked questions about Mixtral 8x7B Instruct v0.1 upscaled i1

How much VRAM does Mixtral 8x7B Instruct v0.1 upscaled i1 need?

About 36.0 GB at IQ3_XS with an 8K context window. That is the 33.6 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 8x7B Instruct v0.1 upscaled i1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 19.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Mixtral 8x7B Instruct v0.1 upscaled i1 GGUF?

It depends on the quant: IQ1_S is 16.7 GB, IQ1_M is 18.6 GB, IQ2_XXS is 21.9 GB, IQ2_S is 24.2 GB, IQ2_XS is 24.3 GB, IQ2_M is 27.0 GB and 11 more.

What context length does Mixtral 8x7B Instruct v0.1 upscaled 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 1.9 GB at 8K and 7.5 GB at 32K.

Where can I download Mixtral 8x7B Instruct v0.1 upscaled i1 GGUF?

The files are on Hugging Face in the mradermacher/Mixtral-8x7B-Instruct-v0.1-upscaled-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 Mixtral 8x7B Instruct v0.1 upscaled 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.

Other GGUFs of Mixtral-8x7B-Instruct-v0.1-upscaled

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Mistral models

All Mistral →

Other large models

All large models →