Upload 2 files

app.py

@@ -0,0 +1,28 @@
+from PIL import Image
+import requests
+import gradio as gr
+
+from transformers import BlipProcessor, BlipForConditionalGeneration
+
+model_id = "Salesforce/blip-image-captioning-base"
+
+model = BlipForConditionalGeneration.from_pretrained(model_id)
+processor = BlipProcessor.from_pretrained(model_id)
+
+def parse_input(input):
+    # this is a weblink
+    if isinstance(input, str):
+        image = Image.open(requests.get(input, stream=True).raw).convert('RGB')
+    # otherwise it's an image
+    else:
+        image = Image.open(input).convert("RGB")
+    
+    return image
+
+def launch(image):
+    inputs = processor(image, return_tensors="pt")
+    out = model.generate(**inputs)
+    return processor.decode(out[0], skip_special_tokens=True)
+
+iface = gr.Interface(launch, inputs="text", outputs="text")
+iface.launch()

requirements.txt

@@ -0,0 +1 @@
+git+https://github.com/younesbelkada/transformers.git@add-blip