XD

xDAN APUS4.0 MoE Initial i1

XdanGGUF by mradermacherUpdated 2024-12-01
Download on Hugging Face
Parameters114B
Context195K
Layers60
Architecturellama
Smallest quant23.4 GB
Quick answer

xDAN APUS4.0 MoE Initial i1 needs about 40.0 GB of memory at IQ2_M (a 37.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 11 quants, from 23.4 GB (IQ1_S) to 49.1 GB (IQ3_S). Longer context needs more: at 32K tokens the KV cache alone adds about 8.1 GB.

xDAN APUS4.0 MoE Initial i1 is a 114B-parameter Xdan model in GGUF format, quantized by mradermacher. Its trained context length is 195K tokens. The repository has 11 quant files, from 23.4 GB (IQ1_S) to 49.1 GB (IQ3_S).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S23.4 GB 25.9 GB
Too big Get file →
IQ1_M25.9 GB 28.4 GB
Too big Get file →
IQ2_XXS30.0 GB 32.5 GB
Too big Get file →
IQ2_XS33.3 GB 35.9 GB
Too big Get file →
IQ2_S34.2 GB 36.8 GB
Too big Get file →
IQ2_M37.5 GB 40.0 GB
Too big Get file →
Q2_K41.5 GB 44.0 GB
Too big Get file →
IQ3_XXS43.7 GB 46.2 GB
Too big Get file →
IQ3_XS46.4 GB 49.0 GB
Too big Get file →
Q3_K_S49.0 GB 51.5 GB
Too big Get file →
IQ3_S49.1 GB 51.6 GB
Too big Get file →

What GPU do you need for xDAN APUS4.0 MoE Initial 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: 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 IQ1_M (25.9 GB file, about 28.4 GB in use).
  • 48 GB VRAM: the best quant that fits is IQ3_XXS (43.7 GB file, about 46.2 GB in use).

Which xDAN APUS4.0 MoE Initial i1 quant should you choose?

IQ2_M (37.5 GB) is the middle option in this repository. 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.

If memory is tight, the smallest file is IQ1_S at 23.4 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
Parameters114B (approximate, mixture of experts)
Layers60
Trained context200,000 tokens
KV cache (fp16)240 KB per token

Source

Quantized bymradermacher
FamilyXdan
Downloads (HF)111
Last updated2024-12-01

Frequently asked questions about xDAN APUS4.0 MoE Initial i1

How much VRAM does xDAN APUS4.0 MoE Initial i1 need?

About 40.0 GB at IQ2_M with an 8K context window. That is the 37.5 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 xDAN APUS4.0 MoE Initial i1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 25.9 GB, so you would need a larger GPU or CPU offloading.

What is the file size of xDAN APUS4.0 MoE Initial i1 GGUF?

It depends on the quant: IQ1_S is 23.4 GB, IQ1_M is 25.9 GB, IQ2_XXS is 30.0 GB, IQ2_XS is 33.3 GB, IQ2_S is 34.2 GB, IQ2_M is 37.5 GB and 5 more.

What context length does xDAN APUS4.0 MoE Initial i1 support?

The GGUF header lists a trained context length of 200,000 tokens (195K). The KV cache grows with the context you use: about 2.0 GB at 8K and 8.1 GB at 32K.

Where can I download xDAN APUS4.0 MoE Initial i1 GGUF?

The files are on Hugging Face in the mradermacher/xDAN-APUS4.0-MoE-Initial-i1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Other GGUFs of xDAN-APUS4.0-MoE-Initial

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Xdan models

All Xdan →

Other large models

All large models →