EV

Evillain StarDust 20B v0.2

EvillainGGUF by RichardErkhovUpdated 2024-06-07
Download on Hugging Face
Parameters20B
Context4K
Layers—
Architecturellama
Smallest quant1.1 GB
Quick answer

Evillain StarDust 20B v0.2 needs about 5.5 GB of memory at Q4_K_M (a 4.4 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 22 quants, from 1.1 GB (Q8_0) to 11.4 GB (IQ4_NL).

Evillain StarDust 20B v0.2 is a 20B-parameter Evillain model in GGUF format, quantized by RichardErkhov. Its trained context length is 4K tokens. The repository has 22 quant files, from 1.1 GB (Q8_0) to 11.4 GB (IQ4_NL).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q8_01.1 GB 1.6 GB
Fits Get file →
Q6_K1.1 GB 1.7 GB
Fits Get file →
Q5_11.4 GB 2.0 GB
Fits Get file →
Q5_K_M1.5 GB 2.1 GB
Fits Get file →
Q5_K1.8 GB 2.4 GB
Fits Get file →
Q5_K_S2.2 GB 2.8 GB
Fits Get file →
Q5_03.2 GB 3.9 GB
Fits Get file →
Q4_13.9 GB 4.7 GB
Fits Get file →
Q4_K_M4.4 GB 5.2 GB
Fits Get file →
Q4_K6.0 GB 6.8 GB
Fits Get file →
Q2_K7.4 GB 8.4 GB
Too big Get file →
Q4_K_S8.1 GB 9.1 GB
Too big Get file →
IQ3_XS8.2 GB 9.2 GB
Too big Get file →
IQ3_S8.7 GB 9.7 GB
Too big Get file →
Q3_K_S8.7 GB 9.7 GB
Too big Get file →
IQ3_M9.2 GB 10.2 GB
Too big Get file →
Q3_K9.7 GB 10.8 GB
Too big Get file →
Q3_K_M9.7 GB 10.8 GB
Too big Get file →
Q3_K_L10.6 GB 11.8 GB
Too big Get file →
IQ4_XS10.7 GB 11.9 GB
Too big Get file →
Q4_011.3 GB 12.5 GB
Too big Get file →
IQ4_NL11.4 GB 12.5 GB
Too big Get file →

What GPU do you need for Evillain StarDust 20B v0.2?

These figures add the file size, an fp16 KV cache at an 8K context window 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: the best quant that fits is Q5_K_S (2.2 GB file, about 3.0 GB in use).
  • 6 GB VRAM: the best quant that fits is Q4_K_M (4.4 GB file, about 5.5 GB in use).
  • 8 GB VRAM: the best quant that fits is Q4_K (6.0 GB file, about 7.2 GB in use).
  • 12 GB VRAM: the best quant that fits is Q3_K (9.7 GB file, about 11.4 GB in use).
  • 16 GB VRAM: the best quant that fits is IQ4_NL (11.4 GB file, about 13.2 GB in use).

Which Evillain StarDust 20B v0.2 quant should you choose?

Q4_K_M is the usual starting point for Evillain StarDust 20B v0.2: it balances file size and quality well. Here it is a 4.4 GB file.

The Q8_0 file is 1.1 GB, about 0.2 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 Q8_0 at 1.1 GB, with the quality trade-offs that come with fewer bits. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters20B
Trained context4,096 tokens

Source

Quantized byRichardErkhov
FamilyEvillain
Downloads (HF)389
Last updated2024-06-07

Frequently asked questions about Evillain StarDust 20B v0.2

How much VRAM does Evillain StarDust 20B v0.2 need?

About 5.5 GB at Q4_K_M with an 8K context window. That is the 4.4 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 Evillain StarDust 20B v0.2 on an 8 GB GPU?

Yes, with Q4_K: the 6.0 GB file needs about 7.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Evillain StarDust 20B v0.2 GGUF?

It depends on the quant: Q8_0 is 1.1 GB, Q6_K is 1.1 GB, Q5_1 is 1.4 GB, Q5_K_M is 1.5 GB, Q5_K is 1.8 GB, Q5_K_S is 2.2 GB and 16 more.

What context length does Evillain StarDust 20B v0.2 support?

The GGUF header lists a trained context length of 4,096 tokens (4K).

Where can I download Evillain StarDust 20B v0.2 GGUF?

The files are on Hugging Face in the RichardErkhov/Evillain_-_StarDust_20B_v0.2-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Similar models

Browse all →

More from RichardErkhov

All by RichardErkhov →

Other medium models

All medium models →