CO

COTHuginn 4.5 19b

CothuginnGGUF by actionpaceUpdated 2023-09-09
Download on Hugging Face
Parameters19.4B
Context4K
Layers60
Architecturellama
Smallest quant8.4 GB
Quick answer

COTHuginn 4.5 19b needs about 22.3 GB of memory at Q4_K_M (a 11.7 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 7 quants, from 8.4 GB (Q3_K_S) to 14.5 GB (Q5_1).

COTHuginn 4.5 19b is a 19.4B-parameter Cothuginn model in GGUF format, quantized by actionpace. Its trained context length is 4K tokens. The repository has 7 quant files, from 8.4 GB (Q3_K_S) to 14.5 GB (Q5_1).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q3_K_S8.4 GB 13.9 GB
Too big Get file →
Q3_K_M9.4 GB 14.9 GB
Too big Get file →
Q3_K_L10.3 GB 15.8 GB
Too big Get file →
Q4_K_S11.0 GB 16.5 GB
Too big Get file →
Q4_K_M11.7 GB 17.2 GB
Too big Get file →
Q4_112.1 GB 17.7 GB
Too big Get file →
Q5_114.5 GB 20.1 GB
Too big Get file →

What GPU do you need for COTHuginn 4.5 19b?

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 Q4_1 (12.1 GB file, about 22.7 GB in use).
  • 32 GB VRAM: the best quant that fits is Q5_1 (14.5 GB file, about 25.1 GB in use).

Which COTHuginn 4.5 19b quant should you choose?

Q4_K_M is the usual starting point for COTHuginn 4.5 19b: it balances file size and quality well. Here it is a 11.7 GB file.

If memory is tight, the smallest file is Q3_K_S at 8.4 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters19.4B
Layers60
Trained context4,096 tokens
KV cache (fp16)1200 KB per token

Source

Quantized byactionpace
FamilyCothuginn
Licenseother
Downloads (HF)32
Last updated2023-09-09

Frequently asked questions about COTHuginn 4.5 19b

How much VRAM does COTHuginn 4.5 19b need?

About 22.3 GB at Q4_K_M with an 8K context window. That is the 11.7 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 COTHuginn 4.5 19b on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 19.0 GB, so you would need a larger GPU or CPU offloading.

What is the file size of COTHuginn 4.5 19b GGUF?

It depends on the quant: Q3_K_S is 8.4 GB, Q3_K_M is 9.4 GB, Q3_K_L is 10.3 GB, Q4_K_S is 11.0 GB, Q4_K_M is 11.7 GB, Q4_1 is 12.1 GB and 1 more.

What context length does COTHuginn 4.5 19b support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 10.1 GB at 8K and 5.0 GB at 4K.

Where can I download COTHuginn 4.5 19b GGUF?

The files are on Hugging Face in the actionpace/COTHuginn-4.5-19b repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does COTHuginn 4.5 19b 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.

Similar models

Browse all →

More from actionpace

All by actionpace →

More Cothuginn models

All Cothuginn →

Other medium models

All medium models →