LE

LewdGem 40B

LewdgemGGUF by Undi95Updated 2024-03-14
Download on Hugging Face
Parameters38.5B
Context4K
Layers40
Architecturellama
Smallest quant21.7 GB
Quick answer

LewdGem 40B needs about 33.7 GB of memory at Q5_0 (a 26.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 3 quants, from 21.7 GB (Q4_0) to 40.9 GB (Q8_0).

LewdGem 40B is a 38.5B-parameter Lewdgem model in GGUF format, quantized by Undi95. Its trained context length is 4K tokens. The repository has 3 quant files, from 21.7 GB (Q4_0) to 40.9 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_021.7 GB 25.6 GB
Too big Get file →
Q5_026.5 GB 30.3 GB
Too big Get file →
Q8_040.9 GB 44.8 GB
Too big Get file →

What GPU do you need for LewdGem 40B?

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: the best quant that fits is Q4_0 (21.7 GB file, about 28.9 GB in use).
  • 48 GB VRAM: the best quant that fits is Q5_0 (26.5 GB file, about 33.7 GB in use).

Which LewdGem 40B quant should you choose?

Q5_0 (26.5 GB) is the middle option in this repository. Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.

If memory is tight, the smallest file is Q4_0 at 21.7 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
Parameters38.5B (approximate, mixture of experts)
Layers40
Trained context4,096 tokens
KV cache (fp16)800 KB per token

Source

Quantized byUndi95
FamilyLewdgem
Downloads (HF)10
Last updated2024-03-14

Frequently asked questions about LewdGem 40B

How much VRAM does LewdGem 40B need?

About 33.7 GB at Q5_0 with an 8K context window. That is the 26.5 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 LewdGem 40B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 28.9 GB, so you would need a larger GPU or CPU offloading.

What is the file size of LewdGem 40B GGUF?

It depends on the quant: Q4_0 is 21.7 GB, Q5_0 is 26.5 GB, Q8_0 is 40.9 GB.

What context length does LewdGem 40B support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 6.7 GB at 8K and 3.4 GB at 4K.

Where can I download LewdGem 40B GGUF?

The files are on Hugging Face in the Undi95/LewdGem-40B-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 LewdGem 40B use?

Hugging Face lists the license as cc-by-nc-4.0. 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 Undi95

All by Undi95 →

More Lewdgem models

All Lewdgem →

Other large models

All large models →