MA

MN Magnum v2.5 18.5B Kto Instruct

MagnumGGUF by bartowskiUpdated 2024-08-25
Download on Hugging Face
Parameters18.5B
Context1000K
Layers63
Architecturellama
Smallest quant6.5 GB
Quick answer

MN Magnum v2.5 18.5B Kto Instruct needs about 13.8 GB of memory at Q4_K_M (a 11.2 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 20 quants, from 6.5 GB (IQ2_M) to 37.0 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.5 GB.

MN Magnum v2.5 18.5B Kto Instruct – a large llama-based instruction model

MN Magnum v2.5 18.5B Kto Instruct is a 18.5‑billion parameter llama architecture model. It supports up to 1,024,000 context tokens and offers multiple quantizations for varied memory budgets. The model is available as GGUF files from the bartowski repository.

Good at

  • Instruction following
  • Large context handling
  • High parameter count

Limits

  • No vision support
  • License not specified
  • Large memory requirement for full precision

Best for

  • Chatbot development
  • Long‑form content generation
  • Research prototyping

Who it’s for. Developers and researchers needing a large instruction‑tuned llama model with flexible quantization options. The model is suitable for environments where GPU memory can be allocated for high‑capacity inference.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_M6.5 GB 9.2 GB
Too big Get file →
Q2_K7.1 GB 9.7 GB
Too big Get file →
Q2_K_L7.8 GB 10.4 GB
Too big Get file →
IQ3_XS7.9 GB 10.5 GB
Too big Get file →
Q3_K_S8.2 GB 10.8 GB
Too big Get file →
IQ3_M8.5 GB 11.1 GB
Too big Get file →
Q3_K_M9.1 GB 11.7 GB
Too big Get file →
Q3_K_L9.8 GB 12.5 GB
Too big Get file →
IQ4_XS10.1 GB 12.7 GB
Too big Get file →
Q3_K_XL10.4 GB 13.0 GB
Too big Get file →
Q4_K_S10.7 GB 13.3 GB
Too big Get file →
Q4_K_M11.2 GB 13.8 GB
Too big Get file →
Q4_K_L11.7 GB 14.3 GB
Too big Get file →
Q5_K_S12.8 GB 15.4 GB
Too big Get file →
Q5_K_M13.2 GB 15.8 GB
Too big Get file →
Q5_K_L13.6 GB 16.2 GB
Too big Get file →
Q6_K15.2 GB 17.8 GB
Too big Get file →
Q6_K_L15.5 GB 18.1 GB
Too big Get file →
Q8_019.7 GB 22.3 GB
Too big Get file →
F1637.0 GB 39.7 GB
Too big Get file →

For 8‑GB VRAM, use IQ2_M (6.54 GB file, 9.2 GB memory at 8k context). For 12‑GB VRAM, Q2_K (7.1 GB file, 9.7 GB memory) or Q2_K_L (7.75 GB file, 10.4 GB memory). For 16‑GB VRAM, Q3_K_M (9.09 GB file, 11.7 GB memory) or Q4_K_M (11.23 GB file, 13.8 GB memory).

What GPU do you need for MN Magnum v2.5 18.5B Kto Instruct?

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: the best quant that fits is Q3_K_M (9.1 GB file, about 11.7 GB in use).
  • 16 GB VRAM: the best quant that fits is Q5_K_M (13.2 GB file, about 15.8 GB in use).
  • 24 GB VRAM: the best quant that fits is Q8_0 (19.7 GB file, about 22.3 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (19.7 GB file, about 22.3 GB in use).
  • 48 GB VRAM: the best quant that fits is F16 (37.0 GB file, about 39.7 GB in use).

Which MN Magnum v2.5 18.5B Kto Instruct quant should you choose?

Q4_K_M is the usual starting point for MN Magnum v2.5 18.5B Kto Instruct: it balances file size and quality well. Here it is a 11.2 GB file.

The Q8_0 file is 19.7 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 IQ2_M at 6.5 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
Parameters18.5B
Layers63
Trained context1,024,000 tokens
KV cache (fp16)252 KB per token

Source

Quantized bybartowski
FamilyMagnum
Downloads (HF)653
Last updated2024-08-25

Frequently asked questions about MN Magnum v2.5 18.5B Kto Instruct

How much VRAM does MN Magnum v2.5 18.5B Kto Instruct need?

About 13.8 GB at Q4_K_M with an 8K context window. That is the 11.2 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 MN Magnum v2.5 18.5B Kto Instruct on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 9.2 GB, so you would need a larger GPU or CPU offloading.

What is the file size of MN Magnum v2.5 18.5B Kto Instruct GGUF?

It depends on the quant: IQ2_M is 6.5 GB, Q2_K is 7.1 GB, Q2_K_L is 7.8 GB, IQ3_XS is 7.9 GB, Q3_K_S is 8.2 GB, IQ3_M is 8.5 GB and 14 more.

What context length does MN Magnum v2.5 18.5B Kto Instruct support?

The GGUF header lists a trained context length of 1,024,000 tokens (1000K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.5 GB at 32K.

Where can I download MN Magnum v2.5 18.5B Kto Instruct GGUF?

The files are on Hugging Face in the bartowski/MN-magnum-v2.5-18.5B-kto-Instruct-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run MN Magnum v2.5 on 8 GB VRAM?

Use the IQ2_M quantization; it requires 9.2 GB memory at 8k context, which is close to 8 GB but may need swapping.

What is the smallest quant file available?

IQ2_M is the smallest at 6.54 GB file size.

Does the model support vision tasks?

No, the model has no vision capabilities.

Similar models

Browse all →

More from bartowski

All by bartowski →

More Magnum models

All Magnum →

Other medium models

All medium models →