Spaces:
Runtime error
Runtime error
Update README.md
Browse files
README.md
CHANGED
|
@@ -10,4 +10,49 @@ pinned: false
|
|
| 10 |
license: mit
|
| 11 |
---
|
| 12 |
|
| 13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
license: mit
|
| 11 |
---
|
| 12 |
|
| 13 |
+
|
| 14 |
+
# 🧠 Dementia-Aware Conversational AI Space
|
| 15 |
+
|
| 16 |
+
This Hugging Face Space (`conversation-response`) is a dementia-support focused conversational model using pretrained Hugging Face pipelines for:
|
| 17 |
+
|
| 18 |
+
- 💬 Dialog generation (DialoGPT)
|
| 19 |
+
- ❤️ Emotion classification (RoBERTa)
|
| 20 |
+
- 🧏 ASR (automatic speech recognition) using Whisper
|
| 21 |
+
- 🗣️ TTS (text-to-speech) using SpeechT5
|
| 22 |
+
|
| 23 |
+
---
|
| 24 |
+
|
| 25 |
+
## 🚀 Features
|
| 26 |
+
|
| 27 |
+
### 🧠 Pre-loaded Pipelines
|
| 28 |
+
- `conversational`: `microsoft/DialoGPT-medium`
|
| 29 |
+
- `emotion`: `j-hartmann/emotion-english-distilroberta-base`
|
| 30 |
+
- `asr`: `openai/whisper-base`
|
| 31 |
+
- `tts`: `microsoft/speecht5_tts`
|
| 32 |
+
|
| 33 |
+
### 🔧 Endpoints
|
| 34 |
+
|
| 35 |
+
#### `/generate` → Chat and Emotion
|
| 36 |
+
```json
|
| 37 |
+
POST /generate
|
| 38 |
+
{
|
| 39 |
+
"message": "I feel lonely today"
|
| 40 |
+
}
|
| 41 |
+
|
| 42 |
+
```
|
| 43 |
+
|
| 44 |
+
```
|
| 45 |
+
/transcribe → Audio Transcription (Whisper)
|
| 46 |
+
|
| 47 |
+
Send audio as multipart/form-data.
|
| 48 |
+
|
| 49 |
+
/speak → Text to Speech (SpeechT5)
|
| 50 |
+
|
| 51 |
+
POST /speak
|
| 52 |
+
|
| 53 |
+
{
|
| 54 |
+
"message": "You're safe with me."
|
| 55 |
+
}
|
| 56 |
+
|
| 57 |
+
Returns: audio/wav stream.
|
| 58 |
+
```
|