Instructions to use buctl/addis-scribe-streaming with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use buctl/addis-scribe-streaming with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("buctl/addis-scribe-streaming") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
File size: 613 Bytes
7680526 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 | {
"model": "addis-scribe-streaming.nemo",
"architecture": "Cache-aware FastConformer (NVIDIA Nemotron 3.5 ASR Streaming 0.6B base), hybrid RNNT+CTC, greedy RNNT decoding",
"sample_rate": 16000,
"language_prompt": "am-ET",
"language_prompt_index": 49,
"default_chunk": "320ms",
"chunks": {
"320ms": {
"att_context_size": [
56,
3
],
"seconds": 0.32
},
"1120ms": {
"att_context_size": [
56,
13
],
"seconds": 1.12
}
},
"decoding": "greedy RNNT",
"output": "plain Amharic text, no punctuation",
"hop_samples": 160,
"n_fft": 512,
"name": "Addis Scribe Streaming"
} |