canary 1b: transcribe.cpp GGUF GGUF conversions of nvidia/canary 1b for use with transcribe.cpp. Ported from upstream commit 1698acf, pinned 2026 05 08. Validated against the NeMo reference at transcribe.cpp commit db53eda on 2026 05 08. Offline multilingual speech to text and translation. A 1B parameter multitask AED with a 24 layer FastConformer encoder and a 24 layer Transformer decoder — the original canary release. Supports automatic speech recognition in English, German, Spanish, and French, and translation between supported pairs. Takes a 16 kHz mono WAV and produces a transcript. Not a streaming model. License: CC BY NC 4.0 (non commercial only) — the only canary variant under a non commercial license. Downloads Quantization Download Size WER (LibriSpeech test clean) : : F32 canary 1b F32.gguf 3.8 GB 1.55% F16 canary 1b F16.gguf 1.9 GB 1.55% Q8 0 canary 1b Q8 0.gguf 1.1 GB 1.55% Q6 K canary 1b Q6 K.gguf 891 MB 1.57% Q5 K M canary 1b Q5 K M.gguf 799 MB 1.57% Q4 K M canary 1b Q4 K M.gguf 696 MB 1.55% WER measured on the full LibriSpeech test clean split (2620 utterances) with greedy decoding and no external LM. F32 reference baseline: 1.55%. NVIDIA's self reported number on t…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy