mlx-community/nemotron-3.5-asr-streaming-0.6b

Same weights, converted vs nvidia/nemotron-3.5-asr-streaming-0.6b. Same weights stored in another dtype (F32 → BF16), plus 245 new tensors.

Tensors changed: MLP 0/96, Norms 0/240

2 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse