Downloads · 30 days
0
JoaoZaokk/distil-whisper-large-v3-ptbr-ggml
distil-whisper-large-v3-ptbr-ggml is a automatic speech recognition model from JoaoZaokk. Use it when you need speech turned into text. It is set up for whisper.cpp. The card lists the license as mit.
GGML conversion of freds0/distil-whisper-large-v3-ptbr (a Whisper checkpoint) for whisper.cpp, in f16 plus q40 / q50 / q80 quantizations.
Downloads · 30 days
0
Access
Public
Updated Sep 12, 2026
Repo size
3.3 GB
Likes
0
Public
Click a slice to open those files.
.bin3.3 GB · 100%
From the Hugging Face model README
GGML conversion of freds0/distil-whisper-large-v3-ptbr (a Whisper checkpoint) for whisper.cpp, in f16 plus q4_0 / q5_0 / q8_0 quantizations.
Source checkpoint: freds0/distil-whisper-large-v3-ptbr by freds0 · License: mit (unchanged; this repo only re-packages the weights)
Engine: Load with whisper.cpp (whisper-cli -m <file>), or any app that embeds it.
| File | Quantization | Size | Note |
|---|---|---|---|
ggml-distil-whisper-large-v3-ptbr-f16.bin | f16 | 1520 MB | |
ggml-distil-whisper-large-v3-ptbr-q4_0.bin | q4_0 | 444 MB | |
ggml-distil-whisper-large-v3-ptbr-q5_0.bin | q5_0 | 538 MB | |
ggml-distil-whisper-large-v3-ptbr-q8_0.bin | q8_0 | 818 MB |
f16 is the lossless conversion; q8_0 is nearly identical in accuracy at ~55 % of the size; q5_0/q5_k are the phone-friendly choice; q4_* is smallest with a small accuracy cost.
Converted from the upstream checkpoint with the engine's own converter, then quantized with the engine's quantizer. Each variant was checked by transcribing short Portuguese and English samples before upload.
Weights are derivative works of the upstream model and keep its license. Please cite the original authors (freds0). The conversion and hosting here are maintained by JoaoZaokk so that the download links used by the Odysseus / Open WebUI native apps stay stable. No warranty.