Phi 3 mini 128k instruct needs about 7.5 GB of memory at Q4_K_M (a 3.0 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 7 quants, from 3.0 GB (Q4_K_M) to 18.9 GB (F32). Longer context needs more: at 32K tokens the KV cache alone adds about 16.1 GB.
Phi 3 mini 128k instruct is a 4.73B-parameter Phi model in GGUF format, quantized by rubra-ai. Its trained context length is 128K tokens. The repository has 7 quant files, from 3.0 GB (Q4_K_M) to 18.9 GB (F32).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 3.0 GB | 7.5 GB | Tight | Get file → | |
| Q5_K_M | 3.5 GB | 8.0 GB | Too big | Get file → | |
| Q6_K | 3.9 GB | 8.4 GB | Too big | Get file → | |
| Q8_0 | 5.0 GB | 9.6 GB | Too big | Get file → | |
| BF16 | 9.5 GB | 14.0 GB | Too big | Get file → | |
| F16 | 9.5 GB | 14.0 GB | Too big | Get file → | |
| F32 | 18.9 GB | 23.4 GB | Too big | Get file → |
What GPU do you need for Phi 3 mini 128k instruct?
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: the best quant that fits is Q5_K_M (3.5 GB file, about 8.0 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (5.0 GB file, about 9.6 GB in use).
- 16 GB VRAM: the best quant that fits is BF16 (9.5 GB file, about 14.0 GB in use).
- 24 GB VRAM: the best quant that fits is F32 (18.9 GB file, about 23.4 GB in use).
Which Phi 3 mini 128k instruct quant should you choose?
Q4_K_M is the usual starting point for Phi 3 mini 128k instruct: it balances file size and quality well. Here it is a 3.0 GB file.
The Q8_0 file is 5.0 GB, about 1.7 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.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Phi 3 mini 128k instruct
How much VRAM does Phi 3 mini 128k instruct need?
About 7.5 GB at Q4_K_M with an 8K context window. That is the 3.0 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 Phi 3 mini 128k instruct on an 8 GB GPU?
Yes, with Q5_K_M: the 3.5 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Phi 3 mini 128k instruct GGUF?
It depends on the quant: Q4_K_M is 3.0 GB, Q5_K_M is 3.5 GB, Q6_K is 3.9 GB, Q8_0 is 5.0 GB, BF16 is 9.5 GB, F16 is 9.5 GB and 1 more.
What context length does Phi 3 mini 128k instruct support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 4.0 GB at 8K and 16.1 GB at 32K.
Where can I download Phi 3 mini 128k instruct GGUF?
The files are on Hugging Face in the rubra-ai/Phi-3-mini-128k-instruct-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 Phi 3 mini 128k instruct 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.