Nemotron 3.5 ASR Streaming — tokenizer supplement

Companion files for CoreML deployments of nvidia/nemotron-3.5-asr-streaming-0.6b, extracted from the official .nemo archive:

tokenizer.model — SentencePiece tokenizer model
vocab.txt — WordPiece-style vocabulary listing

These are required for token-level features (e.g. word boosting / context biasing) that need the real tokenizer rather than the decoded vocabulary. Model weights are not included; see e.g. aufklarer/Nemotron-3.5-ASR-Streaming-0.6B-CoreML-INT8 for converted weights. All artifacts remain subject to the NVIDIA license above.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support