MB Zephyria 45b needs about 30.9 GB of memory at Q4_K_M (a 26.7 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 14 quants, from 16.5 GB (Q2_K) to 47.3 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 15.2 GB.
MB Zephyria 45b is a 44.5B-parameter Zephyria model in GGUF format, quantized by mradermacher. Its trained context length is 128K tokens. The repository has 14 quant files, from 16.5 GB (Q2_K) to 47.3 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 16.5 GB | 20.7 GB | Too big | Get file → | |
| IQ3_XS | 18.3 GB | 22.6 GB | Too big | Get file → | |
| Q3_K_S | 19.2 GB | 23.5 GB | Too big | Get file → | |
| IQ3_S | 19.3 GB | 23.6 GB | Too big | Get file → | |
| IQ3_M | 20.0 GB | 24.3 GB | Too big | Get file → | |
| Q3_K_M | 21.5 GB | 25.8 GB | Too big | Get file → | |
| Q3_K_L | 23.4 GB | 27.7 GB | Too big | Get file → | |
| IQ4_XS | 24.0 GB | 28.3 GB | Too big | Get file → | |
| Q4_K_S | 25.3 GB | 29.5 GB | Too big | Get file → | |
| Q4_K_M | 26.7 GB | 30.9 GB | Too big | Get file → | |
| Q5_K_S | 30.6 GB | 34.9 GB | Too big | Get file → | |
| Q5_K_M | 31.4 GB | 35.7 GB | Too big | Get file → | |
| Q6_K | 36.5 GB | 40.8 GB | Too big | Get file → | |
| Q8_0 | 47.3 GB | 51.6 GB | Too big | Get file → |
What GPU do you need for MB Zephyria 45b?
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 IQ3_S (19.3 GB file, about 23.6 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_K_M (26.7 GB file, about 30.9 GB in use).
- 48 GB VRAM: the best quant that fits is Q6_K (36.5 GB file, about 40.8 GB in use).
Which MB Zephyria 45b quant should you choose?
Q4_K_M is the usual starting point for MB Zephyria 45b: it balances file size and quality well. Here it is a 26.7 GB file.
The Q8_0 file is 47.3 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 Q2_K at 16.5 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
Source
Frequently asked questions about MB Zephyria 45b
How much VRAM does MB Zephyria 45b need?
About 30.9 GB at Q4_K_M with an 8K context window. That is the 26.7 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 MB Zephyria 45b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 20.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of MB Zephyria 45b GGUF?
It depends on the quant: Q2_K is 16.5 GB, IQ3_XS is 18.3 GB, Q3_K_S is 19.2 GB, IQ3_S is 19.3 GB, IQ3_M is 20.0 GB, Q3_K_M is 21.5 GB and 8 more.
What context length does MB Zephyria 45b support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 3.8 GB at 8K and 15.2 GB at 32K.
Where can I download MB Zephyria 45b GGUF?
The files are on Hugging Face in the mradermacher/MB-Zephyria-45b-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.