FI

FireRedPunc

FireredpuncGGUF by cstrUpdated 2026-08-02
Download on Hugging Face
Parameters102M
Context—
Layers—
Architecturefireredpunc
Smallest quant0.1 GB
Quick answer

FireRedPunc needs about 0.6 GB of memory at Q4_K (a 0.1 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 5 quants, from 0.1 GB (IQ4_XS) to 0.2 GB.

FireRedPunc: BERT-based punctuation prediction for Chinese and English

FireRedPunc is a 102M parameter BERT-base model that adds punctuation to ASR output. It supports Chinese and English and is released under Apache 2.0. The model is available in several GGUF quantizations.

Good at

  • Adds punctuation to ASR output
  • Supports Chinese and English
  • Compatible with CrispASR

Limits

  • No vision support
  • Requires 8 k context for full sequence
  • Only 512‑token max sequence length

Best for

  • Post‑processing ASR transcripts
  • Enhancing readability of speech-to-text
  • Integrating into multilingual ASR pipelines

Who it’s for. Developers building ASR systems that need accurate punctuation for Chinese and English. The model is suitable for applications requiring high‑quality text output from speech recognition.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ4_XS0.1 GB 0.6 GB
Fits Get file →
Q4_K0.1 GB 0.6 GB
Fits Get file →
Q4_K0.1 GB 0.6 GB
Fits Get file →
Q8_00.1 GB 0.6 GB
Fits Get file →
—0.2 GB 0.7 GB
Fits Get file →

For 8 GB VRAM, use Q8_0 (0.11 GB file, 0.6 GB memory). If VRAM is limited to 4 GB, IQ4_XS (0.05 GB file, 0.6 GB memory) is a good trade‑off. The unknown quant uses 0.2 GB file and 0.7 GB memory, suitable for larger VRAM budgets.

What GPU do you need for FireRedPunc?

These figures add the file size, an fp16 KV cache at an 8K context window 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: the best quant that fits is the main file (0.2 GB file, about 0.7 GB in use).

Which FireRedPunc quant should you choose?

Q4_K (0.1 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is IQ4_XS at 0.1 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

Architecturefireredpunc
Parameters102M

Source

Quantized bycstr
Licenseapache-2.0
Downloads (HF)5,808
Last updated2026-08-02

Frequently asked questions about FireRedPunc

How much VRAM does FireRedPunc need?

About 0.6 GB at Q4_K with an 8K context window. That is the 0.1 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 FireRedPunc on an 8 GB GPU?

Yes, with the smaller files: the 0.2 GB file needs about 0.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of FireRedPunc GGUF?

It depends on the quant: IQ4_XS is 0.1 GB, Q4_K is 0.1 GB, Q4_K is 0.1 GB, Q8_0 is 0.1 GB, file is 0.2 GB.

Where can I download FireRedPunc GGUF?

The files are on Hugging Face in the cstr/fireredpunc-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 FireRedPunc 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.

Can I run FireRedPunc on 8 GB VRAM?

Yes, the Q8_0 quant uses 0.6 GB memory at 8 k context, which fits in 8 GB VRAM.

Does FireRedPunc support other languages?

It supports Chinese and English only; no other languages are included.

Is the model licensed for commercial use?

The model is released under Apache 2.0, which permits commercial use with attribution.

Similar models

Browse all →

More from cstr

All by cstr →

Other tiny models

All tiny models →