AE

aegolius acadicus v1 30b i1

AegoliusGGUF by mradermacherUpdated 2024-05-06
Download on Hugging Face
Parameters29.8B
Context32K
Layers32
Architecturellama
Smallest quant6.1 GB
Quick answer

aegolius acadicus v1 30b i1 needs about 19.6 GB of memory at Q4_K_M (a 18.0 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 21 quants, from 6.1 GB (IQ1_S) to 24.4 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.

Aegolius Acadicus v1 30B – a 29.8B parameter MOE model

Aegolius Acadicus v1 30B is a 29.8B parameter mixture of experts model based on Llama 2 and Mistral architectures. It was trained with a context length of 32768 tokens and is licensed under llama2. The model is designed for knowledge‑centric tasks.

Good at

  • Knowledge‑heavy queries
  • Multi‑disciplinary reasoning
  • Fine‑tuned Llama2 and Mistral content

Limits

  • Requires 8k context for memory estimates
  • No vision support
  • Large VRAM needed for higher‑quality quants

Best for

  • Research assistants
  • Academic writing aids
  • Technical documentation generators

Who it’s for. Developers and researchers who need a versatile, knowledge‑rich language model that can be run on moderate‑tier hardware with the right quantization. The model is suitable for users who prefer a mixture‑of‑experts approach to cover multiple domains.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S6.1 GB 7.7 GB
Tight Get file →
IQ1_M6.8 GB 8.4 GB
Too big Get file →
IQ2_XXS7.9 GB 9.4 GB
Too big Get file →
IQ2_XS8.8 GB 10.3 GB
Too big Get file →
IQ2_S9.0 GB 10.5 GB
Too big Get file →
IQ2_M9.8 GB 11.4 GB
Too big Get file →
Q2_K10.9 GB 12.5 GB
Too big Get file →
IQ3_XXS11.5 GB 13.0 GB
Too big Get file →
IQ3_XS12.2 GB 13.8 GB
Too big Get file →
Q3_K_S12.9 GB 14.4 GB
Too big Get file →
IQ3_S12.9 GB 14.4 GB
Too big Get file →
IQ3_M13.1 GB 14.7 GB
Too big Get file →
Q3_K_M14.3 GB 15.8 GB
Too big Get file →
Q3_K_L15.5 GB 17.0 GB
Too big Get file →
IQ4_XS15.9 GB 17.5 GB
Too big Get file →
Q4_016.9 GB 18.4 GB
Too big Get file →
Q4_K_S16.9 GB 18.5 GB
Too big Get file →
Q4_K_M18.0 GB 19.6 GB
Too big Get file →
Q5_K_S20.5 GB 22.1 GB
Too big Get file →
Q5_K_M21.1 GB 22.7 GB
Too big Get file →
Q6_K24.4 GB 26.0 GB
Too big Get file →

For 8‑GB VRAM, use IQ1_S (6.14 GB file, 7.7 GB memory) or IQ1_M (6.79 GB file, 8.4 GB memory). If you have 12‑GB VRAM, IQ2_S (8.96 GB file, 10.5 GB memory) or Q2_K (10.88 GB file, 12.5 GB memory) are suitable. Higher‑quality quants such as Q5_K_M require 21.14 GB file and 22.7 GB memory, so they need larger GPUs.

What GPU do you need for aegolius acadicus v1 30b i1?

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 IQ1_S (6.1 GB file, about 7.7 GB in use).
  • 12 GB VRAM: the best quant that fits is IQ2_M (9.8 GB file, about 11.4 GB in use).
  • 16 GB VRAM: the best quant that fits is Q3_K_M (14.3 GB file, about 15.8 GB in use).
  • 24 GB VRAM: the best quant that fits is Q5_K_M (21.1 GB file, about 22.7 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (24.4 GB file, about 26.0 GB in use).

Which aegolius acadicus v1 30b i1 quant should you choose?

Q4_K_M is the usual starting point for aegolius acadicus v1 30b i1: it balances file size and quality well. Here it is a 18.0 GB file.

If memory is tight, the smallest file is IQ1_S at 6.1 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters29.8B (approximate, mixture of experts)
Layers32
Trained context32,768 tokens
KV cache (fp16)128 KB per token

Source

Quantized bymradermacher
FamilyAegolius
Licensellama2
Downloads (HF)1,831
Last updated2024-05-06

Frequently asked questions about aegolius acadicus v1 30b i1

How much VRAM does aegolius acadicus v1 30b i1 need?

About 19.6 GB at Q4_K_M with an 8K context window. That is the 18.0 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 aegolius acadicus v1 30b i1 on an 8 GB GPU?

Yes, with IQ1_S: the 6.1 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 aegolius acadicus v1 30b i1 GGUF?

It depends on the quant: IQ1_S is 6.1 GB, IQ1_M is 6.8 GB, IQ2_XXS is 7.9 GB, IQ2_XS is 8.8 GB, IQ2_S is 9.0 GB, IQ2_M is 9.8 GB and 15 more.

What context length does aegolius acadicus v1 30b i1 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.

Where can I download aegolius acadicus v1 30b i1 GGUF?

The files are on Hugging Face in the mradermacher/aegolius-acadicus-v1-30b-i1-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 aegolius acadicus v1 30b i1 use?

Hugging Face lists the license as llama2. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run this model on 8 GB VRAM?

Yes, with IQ1_S or IQ1_M quants the estimated memory at 8k context is below 9 GB.

Does this model support vision?

No, the model has no vision capabilities.

What context length does it support?

It was trained with a context length of 32768 tokens.

Other GGUFs of aegolius-acadicus-v1-30b

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Aegolius models

All Aegolius →

Other medium models

All medium models →