Neuraltranslate 27b Mt Nah Es v1 needs about 21.2 GB of memory at Q4_K_M (a 16.5 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 23 quants, from 6.3 GB (IQ1_S) to 22.2 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 16.6 GB.
Neuraltranslate 27b: Spanish‑to‑Nahuatl translation model
Neuraltranslate 27b is a Gemma 3 27b Instruct model fine‑tuned on Spanish‑Nahuatl data. It supports Spanish and Náhuatl language pairs. The model is released under the MIT license.
Good at
- Spanish to Nahuatl translation
- Low‑resource language support
- Fine‑tuned Gemma 3 architecture
Limits
- No vision support
- Limited to Spanish‑Nahuatl pair
- Large VRAM requirement for full precision
Best for
- Academic research on indigenous languages
- Developers building translation tools
- Language preservation projects
Who it’s for. Researchers and developers needing a dedicated Spanish‑to‑Nahuatl translation model. The model is suitable for projects that require a specialized language pair and can accommodate the memory demands of a 27b parameter model.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_S | 6.3 GB | 10.9 GB | Too big | Get file → | |
| IQ1_M | 6.8 GB | 11.5 GB | Too big | Get file → | |
| IQ2_XXS | 7.7 GB | 12.3 GB | Too big | Get file → | |
| IQ2_XS | 8.4 GB | 13.1 GB | Too big | Get file → | |
| IQ2_S | 8.8 GB | 13.4 GB | Too big | Get file → | |
| IQ2_M | 9.5 GB | 14.2 GB | Too big | Get file → | |
| Q2_K_S | 9.8 GB | 14.4 GB | Too big | Get file → | |
| Q2_K | 10.5 GB | 15.2 GB | Too big | Get file → | |
| IQ3_XXS | 10.7 GB | 15.4 GB | Too big | Get file → | |
| IQ3_XS | 11.6 GB | 16.2 GB | Too big | Get file → | |
| IQ3_S | 12.2 GB | 16.8 GB | Too big | Get file → | |
| Q3_K_S | 12.2 GB | 16.8 GB | Too big | Get file → | |
| IQ3_M | 12.5 GB | 17.2 GB | Too big | Get file → | |
| Q3_K_M | 13.4 GB | 18.1 GB | Too big | Get file → | |
| Q3_K_L | 14.5 GB | 19.2 GB | Too big | Get file → | |
| IQ4_XS | 14.8 GB | 19.4 GB | Too big | Get file → | |
| Q4_0 | 15.6 GB | 20.3 GB | Too big | Get file → | |
| Q4_K_S | 15.7 GB | 20.3 GB | Too big | Get file → | |
| Q4_K_M | 16.5 GB | 21.2 GB | Too big | Get file → | |
| Q4_1 | 17.2 GB | 21.8 GB | Too big | Get file → | |
| Q5_K_S | 18.8 GB | 23.4 GB | Too big | Get file → | |
| Q5_K_M | 19.3 GB | 23.9 GB | Too big | Get file → | |
| Q6_K | 22.2 GB | 26.8 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_S (6.26 GB) or IQ1_M (6.8 GB). For 12 GB VRAM, IQ2_XS (8.44 GB) or IQ2_S (8.78 GB) are suitable. For 16 GB VRAM, IQ3_S (12.17 GB) or Q3_K_S (12.17 GB) provide a balance of size and performance. Memory estimates are approximate.
What GPU do you need for Neuraltranslate 27b Mt Nah Es v1?
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 IQ1_M (6.8 GB file, about 11.5 GB in use).
- 16 GB VRAM: the best quant that fits is IQ3_XXS (10.7 GB file, about 15.4 GB in use).
- 24 GB VRAM: the best quant that fits is Q5_K_M (19.3 GB file, about 23.9 GB in use).
- 32 GB VRAM: the best quant that fits is Q6_K (22.2 GB file, about 26.8 GB in use).
Which Neuraltranslate 27b Mt Nah Es v1 quant should you choose?
Q4_K_M is the usual starting point for Neuraltranslate 27b Mt Nah Es v1: it balances file size and quality well. Here it is a 16.5 GB file.
If memory is tight, the smallest file is IQ1_S at 6.3 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 Neuraltranslate 27b Mt Nah Es v1
How much VRAM does Neuraltranslate 27b Mt Nah Es v1 need?
About 21.2 GB at Q4_K_M with an 8K context window. That is the 16.5 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 Neuraltranslate 27b Mt Nah Es v1 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 10.9 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Neuraltranslate 27b Mt Nah Es v1 GGUF?
It depends on the quant: IQ1_S is 6.3 GB, IQ1_M is 6.8 GB, IQ2_XXS is 7.7 GB, IQ2_XS is 8.4 GB, IQ2_S is 8.8 GB, IQ2_M is 9.5 GB and 17 more.
What context length does Neuraltranslate 27b Mt Nah Es v1 support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 4.2 GB at 8K and 16.6 GB at 32K.
Where can I download Neuraltranslate 27b Mt Nah Es v1 GGUF?
The files are on Hugging Face in the mradermacher/neuraltranslate-27b-mt-nah-es-v1-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 Neuraltranslate 27b Mt Nah Es v1 use?
Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Neuraltranslate 27b on 8 GB VRAM?
Yes, use the IQ1_S or IQ1_M quantized files, which are 6.26 GB and 6.8 GB respectively. The estimated VRAM usage is about 10.9 GB and 11.5 GB at 8k context.
Does this model support other language pairs?
No. The model is fine‑tuned specifically for Spanish to Nahuatl translation and is not reliable for other pairs.
What license applies to this model?
The model is released under the MIT license.