PH

phi3 medium 128k

PhiGGUF by nistenUpdated 2024-05-21
Download on Hugging Face
Parameters14B
Context128K
Layers40
Architecturephi3
Smallest quant7.5 GB
Quick answer

phi3 medium 128k needs about 13.7 GB of memory at its main file (a 11.5 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 5 quants, from 7.5 GB to 27.9 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 6.7 GB.

phi3 medium 128k is a 14B-parameter Phi model in GGUF format, quantized by nisten. Its trained context length is 128K tokens. The repository has 5 quant files, from 7.5 GB to 27.9 GB.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—7.5 GB 9.7 GB
Too big Get file →
—10.1 GB 12.3 GB
Too big Get file →
—11.5 GB 13.7 GB
Too big Get file →
—14.8 GB 17.0 GB
Too big Get file →
—27.9 GB 30.1 GB
Too big Get file →

What GPU do you need for phi3 medium 128k?

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 the main file (7.5 GB file, about 9.7 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (11.5 GB file, about 13.7 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (14.8 GB file, about 17.0 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (27.9 GB file, about 30.1 GB in use).

Which phi3 medium 128k quant should you choose?

If memory is tight, the smallest file is the first one in the table at 7.5 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturephi3
Parameters14B
Layers40
Trained context131,072 tokens
KV cache (fp16)200 KB per token

Source

Quantized bynisten
FamilyPhi
Licensemit
Downloads (HF)82
Last updated2024-05-21

Frequently asked questions about phi3 medium 128k

How much VRAM does phi3 medium 128k need?

About 13.7 GB at its main file with an 8K context window. That is the 11.5 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 phi3 medium 128k on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 9.7 GB, so you would need a larger GPU or CPU offloading.

What is the file size of phi3 medium 128k GGUF?

It depends on the quant: file is 7.5 GB, file is 10.1 GB, file is 11.5 GB, file is 14.8 GB, file is 27.9 GB.

What context length does phi3 medium 128k support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.7 GB at 8K and 6.7 GB at 32K.

Where can I download phi3 medium 128k GGUF?

The files are on Hugging Face in the nisten/phi3-medium-128k-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 phi3 medium 128k 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.

Other GGUFs of Phi-3-medium-128k-instruct

Similar models

Browse all →

More from nisten

All by nisten →

More Phi models

All Phi →

Other medium models

All medium models →