TypeWhisper/typewhisper-win model-parakeet-ultra-gguf-v1

Parakeet Ultra 0.6B model file (GGUF Q8_0)

on GitHub

Model file the NVIDIA Parakeet plugin downloads to run Parakeet Ultra 0.6B on the graphics card. This is not an app or plugin release.

moondream/parakeet-ultra, Moondream's retrained nvidia/parakeet-tdt-0.6b-v3, converted to GGUF with scripts/convert-parakeet.py of transcribe.cpp 0.3.1 and quantized to Q8_0 with its transcribe-quantize --quant Q8_0. The conversion started from the same NeMo checkpoint as the int8 export in model-parakeet-ultra-int8-v1: Olicorne/parakeet-tdt-0.6b-v3-ultra-onnx, parakeet-tdt-0.6b-v3-ultra.nemo, SHA-256 91b81b3c09a335baa53208c6c70c97aed1c1165df71a51057f26232de33135da. Tensor names, shapes and quantization types match the published Q8_0 GGUF of parakeet-tdt-0.6b-v3.

License: CC-BY-4.0. Model by Moondream, derived from NVIDIA's parakeet-tdt-0.6b-v3.

SHA-256:

• parakeet-ultra-Q8_0.gguf b8752062d51d5d24f6e1355a0b5973d9a47f0d9cd6db8c29385c191ff6f5dcd9 (739,508,704 bytes)