danbev/parakeet-GGUF

Parakeet TDT 0.6B v3 (GGUF)

GGUF conversions of nvidia/parakeet-tdt-0.6b-v3 for use with whisper.cpp.

Available files

• ggml-parakeet-tdt-0.6b-v3-f32.bin — Full precision (F32)
• ggml-parakeet-tdt-0.6b-v3-f16.bin — Half precision (F16)
• ggml-parakeet-tdt-0.6b-v3-q8_0.bin — 8-bit quantized (Q8_0)
• ggml-parakeet-tdt-0.6b-v3-q4_0.bin — 4-bit quantized (Q4_0)
• ggml-parakeet-tdt-0.6b-v3-q4_k.bin — 4-bit K-quantized (Q4_k)

Usage

Build parakeet-cli:

git clone https://github.com/ggml-org/whisper.cpp.git
cd whisper.cpp
cmake -B build -S .
cmake --build build --target parakeet-cli -j $(nproc)

Download a model (e.g. Q8_0):

hf download danbev/parakeet-GGUF ggml-parakeet-tdt-0.6b-v3-q8_0.bin --local-dir models

Run:

./build/bin/parakeet-cli -m models/ggml-parakeet-tdt-0.6b-v3-q8_0.bin -f samples/jfk.wav