Spaces:
Sleeping
Sleeping
File size: 1,836 Bytes
3bf3346 4ffa42b 7ec684a 063bf3b 7ec684a 33e4eda 7ff30bb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 |
import gradio as gr
from sentence_transformers import CrossEncoder
import torch
import requests
# -------------------------------
# CONFIG
# -------------------------------
HF_MODEL = "cross-encoder/ms-marco-MiniLM-L-12-v2"
JINA_MODEL = "jina-reranker-m0"
JINA_API_KEY = "jina_4075150fa702471c85ddea0a9ad4b306ouE7ymhrCpvxTxX3mScUv5LLDPKQ"
JINA_ENDPOINT = "https://api.jina.ai/v1/rerank"
# -------------------------------
# Load Hugging Face CrossEncoder
# -------------------------------
hf_model = CrossEncoder(HF_MODEL)
def compare_models(query, doc):
# Hugging Face score
raw_score = hf_model.predict([(query, doc)])[0]
hf_score = torch.sigmoid(torch.tensor(raw_score)).item()
# Jina reranker score
headers = {
"Authorization": f"Bearer {JINA_API_KEY}",
"Content-Type": "application/json",
}
payload = {
"model": JINA_MODEL,
"query": query,
"documents": [doc],
}
try:
r = requests.post(JINA_ENDPOINT, headers=headers, json=payload, timeout=20)
r.raise_for_status()
jina_score = r.json()["results"][0]["relevance_score"]
except Exception as e:
jina_score = f"Error: {str(e)}"
return f"Hugging Face ({HF_MODEL}): {round(hf_score,4)}\nJina ({JINA_MODEL}): {jina_score}"
# -------------------------------
# Simple Lite UI
# -------------------------------
with gr.Blocks() as demo:
gr.Markdown("### 🔎 Query vs Document Similarity (HF vs Jina)")
query = gr.Textbox(label="Query", lines=3, placeholder="Paste your query here...")
doc = gr.Textbox(label="Document Chunk", lines=6, placeholder="Paste your document chunk here...")
out = gr.Textbox(label="Scores", lines=3)
btn = gr.Button("Compute Similarity 🚀")
btn.click(compare_models, inputs=[query, doc], outputs=out)
demo.launch()
|