File size: 3,392 Bytes
a0562b3 6bde7ff a0562b3 6bde7ff a0562b3 6bde7ff d682b63 a0562b3 6bde7ff d682b63 126f7c1 d682b63 1785474 d682b63 a0562b3 d682b63 a0562b3 6bde7ff a0562b3 d682b63 a0562b3 d682b63 a0562b3 d682b63 a0562b3 |
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 57 58 59 60 61 62 63 64 65 66 67 68 69 |
# https://www.gradio.app/guides/sharing-your-app#mounting-within-another-fast-api-app
import gradio as gr
import json
import logging
from PredictService import PredictService
log_format = "[%(asctime)s] [%(levelname)s] [%(name)s] %(message)s"
logging.basicConfig(level=logging.INFO, format=log_format)
logger = logging.getLogger()
svc = PredictService()
def api_classification(url):
url_to_use = url
if url_to_use == "exemple":
url_to_use = "https://images.pexels.com/photos/326900/pexels-photo-326900.jpeg?cs=srgb&dl=pexels-pixabay-326900.jpg&fm=jpg"
predictions = svc.predict(url_to_use)
return json.dumps(predictions), url_to_use
with gr.Blocks() as app:
with gr.Tab("BioCLIP API"):
with gr.Row():
with gr.Column():
# https://www.gradio.app/guides/key-component-concepts
gr.HTML(value="""
This <a href="https://huggingface.co/spaces/3oly/grBird">Gradio BioCLIP API</a> is a <a href="https://github.com/Imageomics/bioclip">BioCLIP</a> based prediction.<br/><br/>
<small>origin is a <a href="https://huggingface.co/spaces/imageomics/bioclip-demo">BioCLIP DEMO</a>
<a href="https://huggingface.co/spaces/imageomics/bioclip-demo/discussions/7">discussion</a>
following <a href="https://github.com/boly38/pyBird">pyBird</a> MVP.<br/>
This endpoint is used by <a href="https://bsky.app/profile/botensky.bsky.social">@botEnSky</a> BlueSky bot.</small>
<br/>
<h2>How to use Gradio UI ?</h2>
<ul>
<li>You must input a <b>public url of an image</b> and submit by clicking on "predict".</li>
<li>You will get TreeOfLife predictions as result.</li>
</ul>
<h2>How to use Gradio API (Node.js, Python, Bash Curl) ?</h2>
<ul>
<li>You must follow the demo footer link called "Use via API". There is some example provided
(+ <a href="https://github.com/boly38/bioclipBootstrap?tab=readme-ov-file#content-for-3olygrbird">github bootstrap</a>).</li>
</ul>
<h2>Credits</h2>
<ul>
<li>This API endpoint is (only) an augmented fork of the <a href="https://github.com/Imageomics/bioclip">BioCLIP</a> : a wrapper that provide another form of API call.</li>
<li>Research context, model, and contact are available at <a href="https://github.com/Imageomics/bioclip">Imageomics/bioclip</a>.</li>
</ul>
""", show_label=False)
api_input = gr.Textbox(
lines=1,
label="Input a public image url",
show_label=False,
info="Add image url here.",
value="https://natureconservancy-h.assetsadobe.com/is/image/content/dam/tnc/nature/en/photos/d/o/Downy-woodpecker-Matt-Williams.jpg?crop=0%2C39%2C3097%2C2322&wid=820&hei=615&scl=3.776829268292683"
)
api_classification_btn = gr.Button("predict", variant="primary")
api_classification_output_gallery = gr.Image(label="Input image used")
with gr.Column():
api_classification_output_json = gr.JSON(label="This is classification result") # https://www.gradio.app/docs/gradio/json
api_classification_btn.click(
fn=api_classification,
inputs=[api_input],
outputs=[api_classification_output_json, api_classification_output_gallery],
)
app.queue(max_size=20)
app.launch()
|