VinaLlama2-14B-fix needs about 19.5 GB of memory at Q6_K (a 12.3 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 4 quants, from 8.6 GB (Q4_K_S) to 28.3 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 26.8 GB.
VinaLlama2-14B-fix is a 14.2B-parameter Vinallama model in GGUF format, quantized by qnguyen3. Its trained context length is 32K tokens. The repository has 4 quant files, from 8.6 GB (Q4_K_S) to 28.3 GB.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_S | 8.6 GB | 15.8 GB | Too big | Get file → | |
| Q6_K | 12.3 GB | 19.5 GB | Too big | Get file → | |
| Q8_0 | 15.1 GB | 22.3 GB | Too big | Get file → | |
| — | 28.3 GB | 35.6 GB | Too big | Get file → |
What GPU do you need for VinaLlama2-14B-fix?
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: the best quant that fits is Q4_K_S (8.6 GB file, about 15.8 GB in use).
- 24 GB VRAM: the best quant that fits is Q8_0 (15.1 GB file, about 22.3 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (15.1 GB file, about 22.3 GB in use).
- 48 GB VRAM: the best quant that fits is the main file (28.3 GB file, about 35.6 GB in use).
Which VinaLlama2-14B-fix quant should you choose?
Q6_K (12.3 GB) is the middle option in this repository. Very close to the original quality for most uses, at a clearly larger size than 4-bit.
If memory is tight, the smallest file is Q4_K_S at 8.6 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about VinaLlama2-14B-fix
How much VRAM does VinaLlama2-14B-fix need?
About 19.5 GB at Q6_K with an 8K context window. That is the 12.3 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 VinaLlama2-14B-fix on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 15.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of VinaLlama2-14B-fix GGUF?
It depends on the quant: Q4_K_S is 8.6 GB, Q6_K is 12.3 GB, Q8_0 is 15.1 GB, file is 28.3 GB.
What context length does VinaLlama2-14B-fix support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 6.7 GB at 8K and 26.8 GB at 32K.
Where can I download VinaLlama2-14B-fix GGUF?
The files are on Hugging Face in the qnguyen3/14b-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.