HE

Hermyale Stack 90B

HermyaleGGUF by mradermacherUpdated 2024-09-30
Download on Hugging Face
Parameters91.9B
Context128K
Layers105
Architecturellama
Smallest quant19.8 GB
Quick answer

Hermyale Stack 90B needs about 39.6 GB of memory at IQ3_XXS (a 35.6 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 15 quants, from 19.8 GB (IQ1_S) to 49.3 GB (IQ4_XS). Longer context needs more: at 32K tokens the KV cache alone adds about 14.1 GB.

Hermyale Stack 90B – a 91.9B parameter stack merge model

Hermyale Stack 90B is a 91.9B parameter stack merge of Hermes 3 and Euryale 2.2. It is designed for roleplay and creative writing. The model uses a 131072 token context window and is released under a CC‑BY‑NC‑4.0 license.

Good at

  • Creative roleplay
  • Story generation
  • Unique text styles

Limits

  • Large VRAM requirement
  • No vision support
  • Not suited for complex logical tasks

Best for

  • Chatbot roleplay sessions
  • Writing prompts
  • Creative brainstorming

Who it’s for. Users who want a varied, fun model for creative tasks and are willing to manage VRAM usage. It is suitable for those who prefer a stack merge experience over traditional fine‑tuned models.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S19.8 GB 23.8 GB
Too big Get file →
IQ1_M21.6 GB 25.7 GB
Too big Get file →
IQ2_XXS24.7 GB 28.8 GB
Too big Get file →
IQ2_XS27.4 GB 31.4 GB
Too big Get file →
IQ2_S28.8 GB 32.8 GB
Too big Get file →
IQ2_M31.3 GB 35.3 GB
Too big Get file →
Q2_K34.1 GB 38.2 GB
Too big Get file →
IQ3_XXS35.6 GB 39.6 GB
Too big Get file →
IQ3_XS38.0 GB 42.0 GB
Too big Get file →
Q3_K_S39.9 GB 44.0 GB
Too big Get file →
IQ3_S40.0 GB 44.1 GB
Too big Get file →
IQ3_M41.4 GB 45.4 GB
Too big Get file →
Q3_K_M44.4 GB 48.5 GB
Too big Get file →
Q3_K_L48.3 GB 52.4 GB
Too big Get file →
IQ4_XS49.3 GB 53.3 GB
Too big Get file →

For 8 GB VRAM use the IQ1_S quant (file size 19.8 GB, estimated 23.8 GB memory at 8k context). If you have 12 GB VRAM, IQ1_M (21.65 GB, 25.7 GB memory) is a good choice. For 16 GB VRAM, IQ2_XXS (24.73 GB, 28.8 GB memory) provides a balance between size and performance.

What GPU do you need for Hermyale Stack 90B?

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 IQ1_S (19.8 GB file, about 23.8 GB in use).
  • 32 GB VRAM: the best quant that fits is IQ2_XS (27.4 GB file, about 31.4 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_M (41.4 GB file, about 45.4 GB in use).

Which Hermyale Stack 90B quant should you choose?

IQ3_XXS (35.6 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

If memory is tight, the smallest file is IQ1_S at 19.8 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
Parameters91.9B
Layers105
Trained context131,072 tokens
KV cache (fp16)420 KB per token

Source

Quantized bymradermacher
FamilyHermyale
Downloads (HF)598
Last updated2024-09-30

Frequently asked questions about Hermyale Stack 90B

How much VRAM does Hermyale Stack 90B need?

About 39.6 GB at IQ3_XXS with an 8K context window. That is the 35.6 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 Hermyale Stack 90B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 23.8 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Hermyale Stack 90B GGUF?

It depends on the quant: IQ1_S is 19.8 GB, IQ1_M is 21.6 GB, IQ2_XXS is 24.7 GB, IQ2_XS is 27.4 GB, IQ2_S is 28.8 GB, IQ2_M is 31.3 GB and 9 more.

What context length does Hermyale Stack 90B support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 3.5 GB at 8K and 14.1 GB at 32K.

Where can I download Hermyale Stack 90B GGUF?

The files are on Hugging Face in the mradermacher/Hermyale-stack-90B-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 Hermyale Stack 90B 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.

Can I run Hermyale Stack 90B on 8 GB VRAM?

The smallest quant, IQ1_S, requires about 23.8 GB of VRAM at 8k context, so 8 GB VRAM is insufficient.

Does this model support vision tasks?

No, the model has no vision capabilities.

What license applies to this model?

The model is released under a CC‑BY‑NC‑4.0 license.

Other GGUFs of Hermyale-stack-90B

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Hermyale models

All Hermyale →

Other large models

All large models →