Support multiple image sources for LVM microservice (#451)

Signed-off-by: lvliang-intel <[email protected]>

comps/cores/mega/gateway.py

@@ -2,10 +2,13 @@
 # SPDX-License-Identifier: Apache-2.0
 
 import base64
+import os
+from io import BytesIO
 
 import requests
 from fastapi import Request
 from fastapi.responses import StreamingResponse
+from PIL import Image
 
 from ..proto.api_protocol import (
     AudioChatCompletionRequest,
@@ -74,6 +77,7 @@ def list_parameter(self):
         pass
 
     def _handle_message(self, messages):
+        images = []
         if isinstance(messages, str):
             prompt = messages
         else:
@@ -104,7 +108,6 @@ def _handle_message(self, messages):
                     raise ValueError(f"Unknown role: {msg_role}")
             if system_prompt:
                 prompt = system_prompt + "\n"
-            images = []
             for role, message in messages_dict.items():
                 if isinstance(message, tuple):
                     text, image_list = message
@@ -113,8 +116,24 @@ def _handle_message(self, messages):
                     else:
                         prompt += role + ":"
                     for img in image_list:
-                        response = requests.get(img)
-                        images.append(base64.b64encode(response.content).decode("utf-8"))
+                        # URL
+                        if img.startswith("http://") or img.startswith("https://"):
+                            response = requests.get(img)
+                            image = Image.open(BytesIO(response.content)).convert("RGBA")
+                            image_bytes = BytesIO()
+                            image.save(image_bytes, format="PNG")
+                            img_b64_str = base64.b64encode(image_bytes.getvalue()).decode()
+                        # Local Path
+                        elif os.path.exists(img):
+                            image = Image.open(img).convert("RGBA")
+                            image_bytes = BytesIO()
+                            image.save(image_bytes, format="PNG")
+                            img_b64_str = base64.b64encode(image_bytes.getvalue()).decode()
+                        # Bytes
+                        else:
+                            img_b64_str = img
+
+                        images.append(img_b64_str)
                 else:
                     if message:
                         prompt += role + ": " + message + "\n"

comps/lvms/lvm_tgi.py

@@ -48,7 +48,7 @@ async def lvm(request: LVMDoc):
         async def stream_generator():
             chat_response = ""
             text_generation = await lvm_client.text_generation(
-                prompt=prompt,
+                prompt=image_prompt,
                 stream=streaming,
                 max_new_tokens=max_new_tokens,
                 repetition_penalty=repetition_penalty,

requirements.txt

@@ -5,6 +5,7 @@ httpx
 opentelemetry-api
 opentelemetry-exporter-otlp
 opentelemetry-sdk
+Pillow
 prometheus-fastapi-instrumentator
 pyyaml
 requests