Update README.md

README.md

@@ -1,177 +1,8 @@
----
-title: FLAN-T5 Hackathon Idea Generator
-emoji: 💡
-colorFrom: blue
-colorTo: purple
-sdk: docker
-app_port: 7860
-pinned: false
----
+# Hackathon Idea Generator API (FastAPI)
 
-# FLAN-T5 Hackathon Idea Generator
+This Space runs a FastAPI backend only.
 
-A FastAPI backend service that uses Google's FLAN-T5-Base model to generate hackathon project ideas and detailed build plans.
+## Run command
 
-## Features
-
-- **Generate Ideas**: Creates 3 client-side hackathon ideas with tech stacks, difficulty levels, and time estimates
-- **Detailed Plans**: Expands ideas into 48-hour build plans with architecture, phases, code snippets, and risk analysis
-- **JSON Output**: Returns structured JSON responses for easy integration
-
-## API Endpoints
-
-### `GET /`
-Health check endpoint
-```json
-{
-  "status": "FLAN-T5 Backend Running",
-  "endpoints": ["/generate", "/details"],
-  "model": "google/flan-t5-base"
-}
-```
-
-### `POST /generate`
-Generate 3 hackathon ideas
-
-**Request Body:**
-```json
-{
-  "custom_prompt": "focus on AI and machine learning"
-}
-```
-
-**Response:**
-```json
-{
-  "ideas": [
-    {
-      "id": 1,
-      "title": "Project Title",
-      "elevator": "Brief pitch",
-      "overview": "Detailed description",
-      "primary_tech_stack": ["React", "Node.js"],
-      "difficulty": "Medium",
-      "time_estimate_hours": 24
-    }
-  ],
-  "best_pick_id": 1,
-  "best_pick_reason": "Explanation"
-}
-```
-
-### `POST /details`
-Get detailed build plan for an idea
-
-**Request Body:**
-```json
-{
-  "idea_id": 1,
-  "idea_title": "Project Title"
-}
-```
-
-**Response:**
-```json
-{
-  "id": 1,
-  "title": "Project Title",
-  "mermaid_architecture": "graph LR; A-->B;",
-  "phases": [...],
-  "critical_code_snippets": [...],
-  "ui_components": [...],
-  "risks_and_mitigations": [...]
-}
-```
-
-## Testing the API
-
-### Using curl
-```bash
-# Health check
-curl https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME/
-
-# Generate ideas
-curl -X POST https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME/generate \
-  -H "Content-Type: application/json" \
-  -d '{"custom_prompt": "focus on web3 and blockchain"}'
-
-# Get details
-curl -X POST https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME/details \
-  -H "Content-Type: application/json" \
-  -d '{"idea_id": 1, "idea_title": "NFT Gallery"}'
-```
-
-### Using Python
-```python
-import requests
-
-BASE_URL = "https://huggingface.co/spaces/YOUR_USERNAME/YOUR_SPACE_NAME"
-
-# Generate ideas
-response = requests.post(
-    f"{BASE_URL}/generate",
-    json={"custom_prompt": "AI-powered tools"}
-)
-print(response.json())
-
-# Get details
-response = requests.post(
-    f"{BASE_URL}/details",
-    json={"idea_id": 1, "idea_title": "AI Code Assistant"}
-)
-print(response.json())
-```
-
-## Model Information
-
-- **Model**: google/flan-t5-base
-- **Size**: ~250M parameters
-- **Type**: Sequence-to-sequence transformer
-- **Training**: Instruction-finetuned on diverse tasks
-
-## Performance Notes
-
-- Model runs on CPU (suitable for Hugging Face free tier)
-- First request may be slow due to model loading (~30-60 seconds)
-- Subsequent requests are faster with warm model
-- Max generation length: 512 tokens
-
-## Limitations
-
-- FLAN-T5-Base is a smaller model and may not always produce perfect JSON
-- Complex prompts might require prompt engineering
-- Response quality depends on prompt clarity
-- Model has no memory between requests
-
-## Local Development
-
-### Prerequisites
-- Python 3.10+
-- pip
-
-### Setup
 ```bash
-# Install dependencies
-pip install -r requirements.txt
-
-# Run the server
-uvicorn main:app --host 0.0.0.0 --port 7860
-```
-
-The API will be available at `http://localhost:7860`
-
-## Docker Deployment
-
-### Build Image
-```bash
-docker build -t flan-t5-hackathon .
-```
-
-### Run Container
-```bash
-docker run -p 7860:7860 flan-t5-hackathon
-```
-
-## License
-
-This project uses the FLAN-T5 model which is released under Apache 2.0 license.
+uvicorn app:app --host 0.0.0.0 --port $PORT