Kyllene 57B v1.0 i1 needs about 38.4 GB of memory at Q4_K_M (a 34.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 22 quants, from 12.8 GB (IQ1_S) to 47.1 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 13.4 GB.
Kyllene 57B v1.0 i1 is a 56.7B-parameter Kyllene model in GGUF format, quantized by mradermacher. Its trained context length is 195K tokens. The repository has 22 quant files, from 12.8 GB (IQ1_S) to 47.1 GB (Q6_K).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_S | 12.8 GB | 16.7 GB | Too big | Get file → | |
| IQ1_M | 13.9 GB | 17.8 GB | Too big | Get file → | |
| IQ2_XXS | 15.8 GB | 19.7 GB | Too big | Get file → | |
| IQ2_XS | 17.5 GB | 21.3 GB | Too big | Get file → | |
| IQ2_S | 18.4 GB | 22.3 GB | Too big | Get file → | |
| IQ2_M | 19.9 GB | 23.8 GB | Too big | Get file → | |
| Q2_K | 21.6 GB | 25.4 GB | Too big | Get file → | |
| IQ3_XXS | 22.5 GB | 26.3 GB | Too big | Get file → | |
| IQ3_XS | 23.9 GB | 27.7 GB | Too big | Get file → | |
| Q3_K_S | 25.1 GB | 28.9 GB | Too big | Get file → | |
| IQ3_S | 25.2 GB | 29.0 GB | Too big | Get file → | |
| IQ3_M | 26.1 GB | 30.0 GB | Too big | Get file → | |
| Q3_K_M | 27.9 GB | 31.8 GB | Too big | Get file → | |
| Q3_K_L | 30.4 GB | 34.2 GB | Too big | Get file → | |
| IQ4_XS | 30.9 GB | 34.8 GB | Too big | Get file → | |
| IQ4_NL | 32.7 GB | 36.5 GB | Too big | Get file → | |
| Q4_0 | 32.7 GB | 36.5 GB | Too big | Get file → | |
| Q4_K_S | 32.8 GB | 36.6 GB | Too big | Get file → | |
| Q4_K_M | 34.5 GB | 38.4 GB | Too big | Get file → | |
| Q5_K_S | 39.6 GB | 43.4 GB | Too big | Get file → | |
| Q5_K_M | 40.6 GB | 44.5 GB | Too big | Get file → | |
| Q6_K | 47.1 GB | 50.9 GB | Too big | Get file → |
What GPU do you need for Kyllene 57B v1.0 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: the best quant that fits is IQ2_M (19.9 GB file, about 23.8 GB in use).
- 32 GB VRAM: the best quant that fits is Q3_K_M (27.9 GB file, about 31.8 GB in use).
- 48 GB VRAM: the best quant that fits is Q5_K_M (40.6 GB file, about 44.5 GB in use).
Which Kyllene 57B v1.0 i1 quant should you choose?
Q4_K_M is the usual starting point for Kyllene 57B v1.0 i1: it balances file size and quality well. Here it is a 34.5 GB file.
If memory is tight, the smallest file is IQ1_S at 12.8 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
Source
Frequently asked questions about Kyllene 57B v1.0 i1
How much VRAM does Kyllene 57B v1.0 i1 need?
About 38.4 GB at Q4_K_M with an 8K context window. That is the 34.5 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 Kyllene 57B v1.0 i1 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 16.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Kyllene 57B v1.0 i1 GGUF?
It depends on the quant: IQ1_S is 12.8 GB, IQ1_M is 13.9 GB, IQ2_XXS is 15.8 GB, IQ2_XS is 17.5 GB, IQ2_S is 18.4 GB, IQ2_M is 19.9 GB and 16 more.
What context length does Kyllene 57B v1.0 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 3.4 GB at 8K and 13.4 GB at 32K.
Where can I download Kyllene 57B v1.0 i1 GGUF?
The files are on Hugging Face in the mradermacher/Kyllene-57B-v1.0-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 Kyllene 57B v1.0 i1 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.