{"v":1,"id":"model:hf:handy-computer/granite-speech-4.1-2b-nar-gguf","slug":"model-handy-computer-granite-speech-4-1-2b-nar-gguf","kind":"model","category":"speech","title":"granite-speech-4.1-2b-nar-gguf","summary":"granite-speech-4.1-2b-nar: transcribe.cpp GGUF","source":{"provider":"hf","ref":"handy-computer/granite-speech-4.1-2b-nar-gguf","url":"https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf","rev":"c186f8d7ed6c67b947b90fd8ed9587be971766fa","fetchedAt":"2026-10-02T21:00:13.428Z","etag":"W/\"e01-ZNrUoEtS8vHrQhqA0KsYi60Tb0I\""},"author":{"name":"handy-computer","url":"https://huggingface.co/handy-computer"},"license":{"spdx":"apache-2.0","raw":"apache-2.0","open":true},"metrics":{"downloads":39084,"downloadsWeek":15195,"likes":0,"takenAt":"2026-10-02T21:00:13.428Z"},"tags":["transcribe.cpp","gguf","asr","speech-to-text","granite","granite-speech","nar","non-autoregressive","multilingual","automatic-speech-recognition","en","fr","de","es","pt"],"pipeline":"automatic-speech-recognition","links":{"github":"ggml-org/whisper.cpp","npm":"nodejs-whisper"},"updatedAt":"2026-09-15T07:05:33.000Z","collectedAt":"2026-10-02T21:00:13.428Z","review":{"numbers":["39,084 downloads on Hugging Face","0 likes","license apache-2.0","1.5 GB for granite-speech-4.1-2b-nar-Q4_K_M.gguf","15,195 npm downloads a week for nodejs-whisper","latest nodejs-whisper@0.3.1"],"log":null},"trust":"unlabeled","health":{"status":"alive","checkedAt":"2026-10-02T21:00:13.428Z","http":200},"description":"# granite-speech-4.1-2b-nar: transcribe.cpp GGUF\n\nGGUF conversions of [ibm-granite/granite-speech-4.1-2b-nar](https://huggingface.co/ibm-granite/granite-speech-4.1-2b-nar) for use\nwith [transcribe.cpp](https://github.com/handy-computer/transcribe.cpp).\n\nPorted from upstream commit\n[99a4df9](https://huggingface.co/ibm-granite/granite-speech-4.1-2b-nar/commit/99a4df9),\npinned 2026-05-24.\nValidated against the Transformers reference at transcribe.cpp commit\n[c53af2c](https://github.com/handy-computer/transcribe.cpp/tree/c53af2c)\non 2026-05-24.\n\nOffline multilingual speech-to-text in a single non-autoregressive editor\npass. IBM Granite Speech 4.1-2b NAR shares the Conformer audio encoder\nwith the AR Granite-Speech family but pairs it with a custom MLP-with-\nattention projector and the Granite-4.0-1b LLM used as a bidirectional\neditor (causal mask disabled). One forward pass produces logits over the\nfull transcript; CTC decode yields the final text. No token-by-token loop.\nTakes a 16 kHz mono WAV and produces a transcript. English plus French,\nGerman, Spanish, and Portuguese; ASR only (no translation, no\ntimestamps).\n\n## Downloads\n\n| Quantization | Download | Size | WER (LibriSpeech test-clean) |\n| --- | --- | ---: | ---: |\n| BF16 | [granite-speech-4.1-2b-nar-BF16.gguf](https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf/resolve/main/granite-speech-4.1-2b-nar-BF16.gguf) | 4.51 GB | 1.29% |\n| F16 | [granite-speech-4.1-2b-nar-F16.gguf](https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf/resolve/main/granite-speech-4.1-2b-nar-F16.gguf) | 4.52 GB | 1.29% |\n| Q8_0 | [granite-speech-4.1-2b-nar-Q8_0.gguf](https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf/resolve/main/granite-speech-4.1-2b-nar-Q8_0.gguf) | 2.50 GB | 1.29% |\n| Q6_K | [granite-speech-4.1-2b-nar-Q6_K.gguf](https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf/resolve/mai…\n\nSource: https://huggingface.co/handy-computer/granite-speech-4.1-2b-nar-gguf","install":{"kind":"model","hfId":"handy-computer/granite-speech-4.1-2b-nar-gguf","gated":false,"format":"gguf","files":[{"name":"granite-speech-4.1-2b-nar-BF16.gguf","size":4514736032,"quant":"BF16","sha256":"742575c3cea9ff717ad39b18cf635da7c0453f6238fc01baa687beae223ad7e5"},{"name":"granite-speech-4.1-2b-nar-F16.gguf","size":4515792768,"quant":"F16","sha256":"433c4925b3c7fdfea037a343eed60e4c9655acc25d293131ec6f71fd42dfe93f"},{"name":"granite-speech-4.1-2b-nar-Q4_K_M.gguf","size":1560008832,"quant":"Q4_K_M","sha256":"68537a51d4d6d5a6be6e50f735a08af16646afba70fe8ab80dfe60ba7623325f"},{"name":"granite-speech-4.1-2b-nar-Q5_K_M.gguf","size":1782089344,"quant":"Q5_K_M","sha256":"88d7c7b5b8b59c95bb6580a1e7d5d81cae63943cef71405ff477527b0bb69fca"},{"name":"granite-speech-4.1-2b-nar-Q6_K.gguf","size":1977417568,"quant":"Q6_K","sha256":"cfdcdf223a8132ab58bd983df43e277d84056f2a4bf2115fdbc1d40d387f42e1"},{"name":"granite-speech-4.1-2b-nar-Q8_0.gguf","size":2498105472,"quant":"Q8_0","sha256":"4762df797b9efa1ce0d73d2f5128517c86fdfdba2e9d00cd5cca97e624760b21"}],"totalBytes":16848150016,"suggestedFile":"granite-speech-4.1-2b-nar-Q4_K_M.gguf","requirements":{"ramGb":3,"diskBytes":1560008832,"note":"estimate: suggested file size × 1.15 + 0.5 GB; a real measurement comes with lsh models"},"runWith":["whisper.cpp"],"command":"lsh models install hf:handy-computer/granite-speech-4.1-2b-nar-gguf"}}