Nov 1 10 days ago 03:07 Add Silero VAD 6.2.3 Core ML model Oct 29 13 days ago 20:55 Point to the model-named native repos 19:25 Compare speed against NeMo 19:23 Add the Whisper large-v3 comparison with speeds 19:21 Show that WER matches NeMo, with the multilingual results 18:53 Point the card at scriptrs 0.3 and add more speeds 14:21 Update speeds for the FP16 default 14:17 Add native CUDA weights and update the model card Native safetensors for the scriptrs CUDA backend (Parakeet TDT v2 and v3, Silero VAD v5), a card covering every runtime, a LICENSE covering Parakeet and Silero files, and provenance for the native files. Aug 26 20:57 Add parakeet-v3 scriptrs Core ML layout with license and provenance Apr 10 08:30 Upload split decoder and joint model bundle 08:30 Remove stale parakeet-v2/decoder-joint.mlmodelc 07:56 Upload fused decoder-joint model bundle 07:56 Remove stale parakeet-v2/joint-decision.mlmodelc 07:56 Remove stale parakeet-v2/decoder.mlmodelc 07:55 Upload fused decoder-joint model bundle Apr 3 14:07 Add files using upload-large-folder tool 14:07 initial commit 
