TE

Technoculture Mediquad 4x7b

TechnocultureGGUF by RichardErkhovUpdated 2024-09-01
Download on Hugging Face
Parameters20B
Context8K
Layers32
Architecturellama
Smallest quant1.7 GB
Quick answer

Technoculture Mediquad 4x7b needs about 16.7 GB of memory at Q4_K_M (a 11.9 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 22 quants, from 1.7 GB (Q4_K_S) to 21.0 GB (Q8_0).

Technoculture Mediquad 4x7b is a 20B-parameter Technoculture model in GGUF format, quantized by RichardErkhov. Its trained context length is 8K tokens. The repository has 22 quant files, from 1.7 GB (Q4_K_S) to 21.0 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_S1.7 GB 6.5 GB
Fits Get file →
IQ3_S2.5 GB 7.3 GB
Tight Get file →
IQ4_NL2.9 GB 7.7 GB
Tight Get file →
IQ3_XS4.0 GB 8.8 GB
Too big Get file →
IQ3_M5.8 GB 10.6 GB
Too big Get file →
Q2_K7.2 GB 12.0 GB
Too big Get file →
Q3_K_L7.8 GB 12.6 GB
Too big Get file →
Q3_K_S7.9 GB 12.7 GB
Too big Get file →
Q3_K9.5 GB 14.3 GB
Too big Get file →
Q3_K_M9.5 GB 14.3 GB
Too big Get file →
IQ4_XS10.6 GB 15.4 GB
Too big Get file →
Q4_011.1 GB 15.9 GB
Too big Get file →
Q4_K11.9 GB 16.7 GB
Too big Get file →
Q4_K_M11.9 GB 16.7 GB
Too big Get file →
Q4_112.4 GB 17.2 GB
Too big Get file →
Q5_013.6 GB 18.4 GB
Too big Get file →
Q5_K_S13.6 GB 18.4 GB
Too big Get file →
Q5_K14.0 GB 18.8 GB
Too big Get file →
Q5_K_M14.0 GB 18.8 GB
Too big Get file →
Q5_114.8 GB 19.6 GB
Too big Get file →
Q6_K16.2 GB 21.0 GB
Too big Get file →
Q8_021.0 GB 25.8 GB
Too big Get file →

What GPU do you need for Technoculture Mediquad 4x7b?

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 IQ4_NL (2.9 GB file, about 7.7 GB in use).
  • 12 GB VRAM: the best quant that fits is Q2_K (7.2 GB file, about 12.0 GB in use).
  • 16 GB VRAM: the best quant that fits is Q4_0 (11.1 GB file, about 15.9 GB in use).
  • 24 GB VRAM: the best quant that fits is Q6_K (16.2 GB file, about 21.0 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (21.0 GB file, about 25.8 GB in use).

Which Technoculture Mediquad 4x7b quant should you choose?

Q4_K_M is the usual starting point for Technoculture Mediquad 4x7b: it balances file size and quality well. Here it is a 11.9 GB file.

The Q8_0 file is 21.0 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 Q4_K_S at 1.7 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters20B (approximate, mixture of experts)
Layers32
Trained context8,192 tokens
KV cache (fp16)512 KB per token

Source

Quantized byRichardErkhov
Licenseapache-2.0
Downloads (HF)4
Last updated2024-09-01

Frequently asked questions about Technoculture Mediquad 4x7b

How much VRAM does Technoculture Mediquad 4x7b need?

About 16.7 GB at Q4_K_M with an 8K context window. That is the 11.9 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 Technoculture Mediquad 4x7b on an 8 GB GPU?

Yes, with IQ4_NL: the 2.9 GB file needs about 7.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Technoculture Mediquad 4x7b GGUF?

It depends on the quant: Q4_K_S is 1.7 GB, IQ3_S is 2.5 GB, IQ4_NL is 2.9 GB, IQ3_XS is 4.0 GB, IQ3_M is 5.8 GB, Q2_K is 7.2 GB and 16 more.

What context length does Technoculture Mediquad 4x7b support?

The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 4.3 GB at 8K and 4.3 GB at 8K.

Where can I download Technoculture Mediquad 4x7b GGUF?

The files are on Hugging Face in the RichardErkhov/Technoculture_-_Mediquad-4x7b-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 Technoculture Mediquad 4x7b 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 RichardErkhov

All by RichardErkhov →

More Technoculture models

All Technoculture →

Other medium models

All medium models →