RH

Rhea 72b v0.1

RheaGGUF by RichardErkhovUpdated 2024-10-16
Download on Hugging Face
Parameters72.3B
Context32K
Layers80
Architecturellama
Smallest quant3.7 GB
Quick answer

Rhea 72b v0.1 needs about 65.7 GB of memory at Q4_K_M (a 43.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 22 quants, from 3.7 GB (IQ3_XS) to 76.8 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 85.9 GB.

Rhea 72b v0.1 is a 72.3B-parameter Rhea model in GGUF format, quantized by RichardErkhov. Its trained context length is 32K tokens. The repository has 22 quant files, from 3.7 GB (IQ3_XS) to 76.8 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ3_XS3.7 GB 25.6 GB
Too big Get file →
Q2_K27.1 GB 49.1 GB
Too big Get file →
IQ3_S31.6 GB 53.5 GB
Too big Get file →
Q3_K_S31.6 GB 53.5 GB
Too big Get file →
IQ3_M33.3 GB 55.2 GB
Too big Get file →
Q3_K35.3 GB 57.3 GB
Too big Get file →
Q3_K_M35.3 GB 57.3 GB
Too big Get file →
Q3_K_L38.5 GB 60.5 GB
Too big Get file →
IQ4_XS39.1 GB 61.1 GB
Too big Get file →
Q4_041.0 GB (2 parts) 63.0 GB
Too big Get file →
IQ4_NL41.3 GB (2 parts) 63.2 GB
Too big Get file →
Q4_K_S41.3 GB (2 parts) 63.3 GB
Too big Get file →
Q4_K43.8 GB (2 parts) 65.7 GB
Too big Get file →
Q4_K_M43.8 GB (2 parts) 65.7 GB
Too big Get file →
Q4_145.4 GB (2 parts) 67.4 GB
Too big Get file →
Q5_049.9 GB (2 parts) 71.9 GB
Too big Get file →
Q5_K_S49.9 GB (2 parts) 71.9 GB
Too big Get file →
Q5_K51.3 GB (2 parts) 73.3 GB
Too big Get file →
Q5_K_M51.3 GB (2 parts) 73.3 GB
Too big Get file →
Q5_154.3 GB (2 parts) 76.3 GB
Too big Get file →
Q6_K59.3 GB (2 parts) 81.3 GB
Too big Get file →
Q8_076.8 GB (2 parts) 98.8 GB
Too big Get file →

What GPU do you need for Rhea 72b v0.1?

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 IQ3_XS (3.7 GB file, about 25.6 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_XS (3.7 GB file, about 25.6 GB in use).

Which Rhea 72b v0.1 quant should you choose?

Q4_K_M is the usual starting point for Rhea 72b v0.1: it balances file size and quality well. Here it is a 43.8 GB file.

The Q8_0 file is 76.8 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 IQ3_XS at 3.7 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
Parameters72.3B
Layers80
Trained context32,768 tokens
KV cache (fp16)2560 KB per token

Source

Quantized byRichardErkhov
FamilyRhea
Licenseapache-2.0
Downloads (HF)29
Last updated2024-10-16

Frequently asked questions about Rhea 72b v0.1

How much VRAM does Rhea 72b v0.1 need?

About 65.7 GB at Q4_K_M with an 8K context window. That is the 43.8 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 Rhea 72b v0.1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 25.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Rhea 72b v0.1 GGUF?

It depends on the quant: IQ3_XS is 3.7 GB, Q2_K is 27.1 GB, IQ3_S is 31.6 GB, Q3_K_S is 31.6 GB, IQ3_M is 33.3 GB, Q3_K is 35.3 GB and 16 more.

What context length does Rhea 72b v0.1 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 21.5 GB at 8K and 85.9 GB at 32K.

Where can I download Rhea 72b v0.1 GGUF?

The files are on Hugging Face in the RichardErkhov/davidkim205_-_Rhea-72b-v0.1-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 Rhea 72b v0.1 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 Rhea models

All Rhea →

Other large models

All large models →