uzbek-stt-fastconformer-v1.5 – AI Model by zafarrr | AlphaNeural AI

You can deploy this model and start earning money today!

zafarrr/
uzbek-stt-fastconformer-v1.5

like

0

automatic-speech-recognition

nvidia/stt_uz_fastconformer_hybrid_large_pc

Views

No views yet

Uzbek STT FastConformer v1.5

Uzbek automatic speech recognition model based on NVIDIA's FastConformer architecture.

Model Details

Property	Value
Architecture	FastConformer Hybrid (RNNT + CTC)
Parameters	120M
Base Model	nvidia/stt_uz_fastconformer_hybrid_large_pc
License	Apache 2.0

Usage

python

import nemo.collections.asr as nemo_asr
from huggingface_hub import hf_hub_download

model_path = hf_hub_download(
    repo_id="zafarrr/uzbek-stt-fastconformer-v1.5",
    filename="uzbek_fastconformer_v1.5.nemo"
)

model = nemo_asr.models.ASRModel.restore_from(model_path)
model.eval()
transcriptions = model.transcribe(["audio.wav"])
print(transcriptions[0])

Limitations

Optimized for clean speech
Performance may degrade on noisy audio
Uzbek language only