KAT Dev 72B Exp needs about 45.4 GB of memory at IQ4_XS (a 40.1 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 21 quants, from 0.0 GB to 145.4 GB (BF16).
KAT Dev 72B Exp is a 72.7B-parameter Kat model in GGUF format, quantized by Mungert. Its trained context length is 128K tokens. The repository has 21 quant files, from 0.0 GB to 145.4 GB (BF16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.0 GB | 0.5 GB | Fits | Get file → | |
| IQ1_S | 23.6 GB | 26.9 GB | Too big | Get file → | |
| IQ1_M | 25.2 GB | 28.7 GB | Too big | Get file → | |
| IQ2_XXS | 26.3 GB | 30.0 GB | Too big | Get file → | |
| IQ2_XS | 27.9 GB | 31.7 GB | Too big | Get file → | |
| IQ2_S | 28.1 GB | 32.0 GB | Too big | Get file → | |
| IQ2_M | 29.4 GB | 33.5 GB | Too big | Get file → | |
| IQ3_XXS | 32.3 GB | 36.7 GB | Too big | Get file → | |
| IQ3_XS | 33.2 GB | 37.6 GB | Too big | Get file → | |
| IQ3_M | 35.3 GB | 40.0 GB | Too big | Get file → | |
| IQ4_XS | 40.1 GB | 45.4 GB | Too big | Get file → | |
| IQ4_NL | 41.0 GB | 46.4 GB | Too big | Get file → | |
| Q4_0 | 42.2 GB | 47.7 GB | Too big | Get file → | |
| Q4_K_S | 44.3 GB | 50.2 GB | Too big | Get file → | |
| Q4_1 | 45.6 GB | 51.6 GB | Too big | Get file → | |
| Q5_0 | 50.9 GB | 57.5 GB | Too big | Get file → | |
| Q5_1 | 55.3 GB (2 parts) | 62.5 GB | Too big | Get file → | |
| Q6_K_M | 65.0 GB (2 parts) | 73.2 GB | Too big | Get file → | |
| Q8_0 | 77.3 GB (2 parts) | 87.0 GB | Too big | Get file → | |
| F16 | 103.4 GB (3 parts) | 116.3 GB | Too big | Get file → | |
| BF16 | 145.4 GB (4 parts) | 163.4 GB | Too big | Get file → |
What GPU do you need for KAT Dev 72B Exp?
These figures add the file size, an fp16 KV cache at an 8K context window 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: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.0 GB file, about 0.5 GB in use).
- 32 GB VRAM: the best quant that fits is IQ2_S (28.1 GB file, about 32.0 GB in use).
- 48 GB VRAM: the best quant that fits is Q4_0 (42.2 GB file, about 47.7 GB in use).
Which KAT Dev 72B Exp quant should you choose?
IQ4_XS (40.1 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is the first one in the table at 0.0 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about KAT Dev 72B Exp
How much VRAM does KAT Dev 72B Exp need?
About 45.4 GB at IQ4_XS with an 8K context window. That is the 40.1 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 KAT Dev 72B Exp on an 8 GB GPU?
Yes, with the smaller files: the 0.0 GB file needs about 0.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of KAT Dev 72B Exp GGUF?
It depends on the quant: file is 0.0 GB, IQ1_S is 23.6 GB, IQ1_M is 25.2 GB, IQ2_XXS is 26.3 GB, IQ2_XS is 27.9 GB, IQ2_S is 28.1 GB and 15 more.
What context length does KAT Dev 72B Exp support?
The GGUF header lists a trained context length of 131,072 tokens (128K).
Where can I download KAT Dev 72B Exp GGUF?
The files are on Hugging Face in the Mungert/KAT-Dev-72B-Exp-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 KAT Dev 72B Exp 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.