Instructions to use cortex-ai-nova/NOVA-Oracle-4.6-16bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use cortex-ai-nova/NOVA-Oracle-4.6-16bit with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="cortex-ai-nova/NOVA-Oracle-4.6-16bit")# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq processor = AutoProcessor.from_pretrained("cortex-ai-nova/NOVA-Oracle-4.6-16bit") model = AutoModelForSpeechSeq2Seq.from_pretrained("cortex-ai-nova/NOVA-Oracle-4.6-16bit", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
Uploaded finetuned model
NOVA-Oracle-4.6 is an "upgraded" version of the famous Whisper Large V3, which specializes in speech-to-text (or vice versa) transcription and is a universal "Swiss Army knife" for 99 languages. This model was specifically trained (fine-tuned) for specific tasks.
What it does:
Audio transcription: Turns voice into text (podcasts, lectures, calls, videos). Works in 99+ languages and is fine-tuned on 8 datasets and various sources (open license). Understanding difficult conditions: Thanks to the Whisper large-v3 base, it is resistant to background noise, poor microphones, and fast speech.
✨ What the "catch" and main feature is
🤔 Who it is for This model can be for both an ordinary user who needs to translate a video from Spanish to Russian (example), and for developers and Data Scientists.
- Developed by: cortex-ai-nova
- License: apache-2.0
Cortex-AI
- Downloads last month
- -
