vapi-gemma-api / app.py

Commit History

  1. ULTRA SPEED: 8-bit quantization, greedy decoding, 40 tokens, inference_mode

    Ilke Ileri
    c46fe44
  2. Optimize for speed: reduce to 80 tokens, lower sampling params, add timing logs

    Ilke Ileri
    190133f
  3. Improve conversation quality: use full history, increase tokens to 150, better sampling

    Ilke Ileri
    39bb917
  4. Add /v1/chat/completions route for Vapi compatibility

    Ilke Ileri
    2311595
  5. Optimize for speed: max_tokens=50, greedy decoding for real-time voice

    Ilke Ileri
    b9e9889
  6. Add debug endpoint and improve error handling for Vapi troubleshooting

    Ilke Ileri
    88b53d1
  7. Reduce max_tokens to 100 for faster response to prevent Vapi timeout

    Ilke Ileri
    d54cae5

  1. Add streaming support for Vapi compatibility

    Ilke Ileri
    0ccd1fa
  2. Fix: remove global counter causing Space crash

    Ilke Ileri
    c6b30d3
  3. Add request counter to track incoming Vapi requests

    Ilke Ileri
    dcb9815
  4. Fix: properly extract user messages from Vapi conversation history

    Ilke Ileri
    0385f33
  5. Enhance request logging for better Vapi debugging

    Ilke Ileri
    260002e
  6. Add keyword-based sales filter to redirect off-topic questions

    Ilke Ileri
    7242003
  7. Fix: revert system prompt, add full OpenAI-compatible response format

    Ilke Ileri
    a1e3c35
  8. Add system prompt guard to enforce sales-only responses

    Ilke Ileri
    a73c020
  9. Fix CORS issues and add request logging for Vapi integration

    Ilke Ileri
    e48c956
  10. Improve response formatting and generation parameters

    Ilke Ileri
    e683a4c
  11. Add HF_TOKEN support for gated model access

    Ilke Ileri
    25839d0
  12. Add Vapi Gemma API application

    Ilke Ileri
    692ef6b