PO

Poro 34B chat

PoroGGUF by bartowskiUpdated 2024-06-15
Download on Hugging Face
Parameters35.1B
Context7K
Layers54
Architecturebloom
Smallest quant13.4 GB
Quick answer

Poro 34B chat needs about 38.0 GB of memory at Q4_K_M (a 24.8 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 18 quants, from 13.4 GB (IQ2_XS) to 140.5 GB (F32).

Poro 34B chat is a 35.1B-parameter Poro model in GGUF format, quantized by bartowski. Its trained context length is 7K tokens. The repository has 18 quant files, from 13.4 GB (IQ2_XS) to 140.5 GB (F32).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ2_XS13.4 GB 25.0 GB
Too big Get file →
IQ2_S13.9 GB 25.5 GB
Too big Get file →
IQ2_M14.8 GB 26.4 GB
Too big Get file →
Q2_K16.0 GB 27.6 GB
Too big Get file →
IQ3_XXS16.6 GB 28.2 GB
Too big Get file →
IQ3_XS17.6 GB 29.2 GB
Too big Get file →
Q3_K_S18.0 GB 29.6 GB
Too big Get file →
IQ3_M19.6 GB 31.2 GB
Too big Get file →
Q3_K_M21.0 GB 32.6 GB
Too big Get file →
IQ4_XS21.4 GB 33.0 GB
Too big Get file →
Q4_K_S22.6 GB 34.2 GB
Too big Get file →
Q3_K_L22.7 GB 34.3 GB
Too big Get file →
Q4_K_M24.8 GB 36.4 GB
Too big Get file →
Q5_K_S26.6 GB 38.2 GB
Too big Get file →
Q5_K_M28.4 GB 40.0 GB
Too big Get file →
Q6_K31.0 GB 42.6 GB
Too big Get file →
Q8_039.1 GB 50.7 GB
Too big Get file →
F32140.5 GB (4 parts) 152.1 GB
Too big Get file →

What GPU do you need for Poro 34B chat?

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: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is Q3_K_S (18.0 GB file, about 31.2 GB in use).
  • 48 GB VRAM: the best quant that fits is Q6_K (31.0 GB file, about 44.2 GB in use).

Which Poro 34B chat quant should you choose?

Q4_K_M is the usual starting point for Poro 34B chat: it balances file size and quality well. Here it is a 24.8 GB file.

The Q8_0 file is 39.1 GB, about 1.6 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.

If memory is tight, the smallest file is IQ2_XS at 13.4 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

Architecturebloom
Parameters35.1B
Layers54
Trained context7,168 tokens
KV cache (fp16)1512 KB per token

Source

Quantized bybartowski
FamilyPoro
Licenseapache-2.0
Downloads (HF)210
Last updated2024-06-15

Frequently asked questions about Poro 34B chat

How much VRAM does Poro 34B chat need?

About 38.0 GB at Q4_K_M with an 8K context window. That is the 24.8 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 Poro 34B chat on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 26.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Poro 34B chat GGUF?

It depends on the quant: IQ2_XS is 13.4 GB, IQ2_S is 13.9 GB, IQ2_M is 14.8 GB, Q2_K is 16.0 GB, IQ3_XXS is 16.6 GB, IQ3_XS is 17.6 GB and 12 more.

What context length does Poro 34B chat support?

The GGUF header lists a trained context length of 7,168 tokens (7K). The KV cache grows with the context you use: about 12.7 GB at 8K and 11.1 GB at 7K.

Where can I download Poro 34B chat GGUF?

The files are on Hugging Face in the bartowski/Poro-34B-chat-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 Poro 34B chat use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Similar models

Browse all →

More from bartowski

All by bartowski →

More Poro models

All Poro →

Other large models

All large models →