AI Fairness and Bias Mitigation
Emotion Tags
normalslowcryingsleepysighchuckle
Focus on Depression and Anxiety
Guided Emotion and Intonation
Inpatient and Outpatient Use
Input Streaming for Lower Latency
LLM-based Customizability
Mental Health Screening Tool
Model Tokenizer Type
Non-streaming (CNN-based) tokenizer
Open Source Release Planned
Orpheus Speech Models
Medium (3B)Small (1B)Tiny (400M)Nano (150M)
Pretrained and Finetuned Models
Pretrained modelsFinetuned models
Sample Finetuning Scripts
Short Speech Clip Analysis
Sliding Window Detokenizer
Streaming Inference Speed
Faster than playback on A100 40GB for 3B model
Training Data Volume
100k+ hours of speech, billions of text tokens
Unbiased Data Acquisition
Voice Biomarker Detection