Upload 3 files

Dockerfile

@@ -0,0 +1,18 @@
+# Use official Python image as base
+FROM python:3.11-slim
+
+# Set working directory
+WORKDIR /app
+
+# Copy requirements and install dependencies
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy application code
+COPY . .
+
+# Expose port 7860 for Hugging Face Spaces
+EXPOSE 7860
+
+# Command to run the FastAPI app with uvicorn
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

main.py

@@ -0,0 +1,86 @@
+import os
+from fastapi import FastAPI, HTTPException, Request
+from pydantic import BaseModel
+import httpx
+from dotenv import load_dotenv
+
+load_dotenv()
+
+app = FastAPI()
+
+HF_ENDPOINT = os.getenv("HF_ENDPOINT", "https://yv6q340a33xhe4bs.us-east-1.aws.endpoints.huggingface.cloud")
+HF_API_KEY = os.getenv("HF_API_KEY", None)
+
+headers = {
+    "Accept": "application/json",
+    "Content-Type": "application/json",
+}
+
+if HF_API_KEY:
+    headers["Authorization"] = f"Bearer {HF_API_KEY}"
+
+class EmbeddingRequest(BaseModel):
+    inputs: str
+    parameters: dict = {}
+
+class OpenAIEmbeddingData(BaseModel):
+    object: str
+    embedding: list
+    index: int
+
+class OpenAIEmbeddingResponse(BaseModel):
+    object: str
+    data: list
+    model: str
+    usage: dict
+
+@app.post("/v1/embeddings", response_model=OpenAIEmbeddingResponse)
+async def embeddings(request: EmbeddingRequest):
+    payload = {
+        "inputs": request.inputs,
+        "parameters": request.parameters
+    }
+    async with httpx.AsyncClient() as client:
+        try:
+            response = await client.post(HF_ENDPOINT, json=payload, headers=headers)
+            response.raise_for_status()
+        except httpx.HTTPError as e:
+            raise HTTPException(status_code=500, detail=f"Failed to fetch embeddings: {str(e)}")
+
+    hf_data = response.json()
+    # Huggingface response is a list of floats (embedding vector), so hf_data is a list
+    if isinstance(hf_data, list):
+        embeddings = hf_data
+    elif isinstance(hf_data, dict) and "embeddings" in hf_data:
+        embeddings = hf_data["embeddings"]
+    else:
+        embeddings = hf_data  # fallback
+
+    openai_response = OpenAIEmbeddingResponse(
+        object="list",
+        data=[OpenAIEmbeddingData(object="embedding", embedding=embeddings, index=0)],
+        model="Qwen3-Embedding-8B",
+        usage={"prompt_tokens": 0, "total_tokens": 0}
+    )
+
+    return openai_response
+
+@app.get("/health")
+async def health():
+    return {"status": "ok"}
+
+@app.get("/v1/modals")
+async def list_modals():
+    return {
+        "object": "list",
+        "data": [
+            {
+                "id": "Qwen3-Embedding-8B",
+                "object": "model",
+                "owned_by": "qwen",
+                "permission": [],
+                "root": "Qwen3-Embedding-8B",
+                "parent": None
+            }
+        ]
+    }

requirements.txt

@@ -0,0 +1,4 @@
+fastapi
+uvicorn[standard]
+httpx
+python-dotenv