CoolFace
Apppublic

Atishay210/chatbot-api

sourceHugging Facemitupdated 5mo agoView on Hugging Face
0likes
App README

AI Chatbot — FastAPI Backend

Fine-tuned Mistral-7B-Instruct-v0.2 chatbot backend running on HuggingFace Spaces.

Endpoints

MethodPathDescription
GET/healthModel status + VRAM info
POST/session/newCreate conversation session
POST/chatStream response (SSE)
GET/session/{id}/historyGet chat history
DELETE/session/{id}Clear session
GET/docsSwagger UI

Tech stack

  • Model: Mistral-7B-Instruct-v0.2 + QLoRA adapter (atishay210/chatbot-mistral-qlora)
  • Quantization: 4-bit NF4 via bitsandbytes
  • Serving: FastAPI + Uvicorn + Server-Sent Events streaming
  • Hardware: HuggingFace Spaces T4 GPU (16GB VRAM)