Modifications: handy-computer converted
https://huggingface.co/nvidia/nemotron-3.5-asr-streaming-0.6b at commit
24b151a to GGUF for transcribe.cpp
(https://github.com/handy-computer/transcribe.cpp). Conversion repackages the
weights, renaming and re-laying-out tensors; the quantized files (Q8_0, Q6_K,
Q5_K_M, Q4_K_M) also store them at reduced numeric precision. The weights were
not retrained or fine-tuned.