Update app.py

app.py

@@ -6,6 +6,7 @@ from together import Together
 import base64
 from threading import Thread
 import time
+import io
 
 # Initialize Together client
 client = None
@@ -19,11 +20,14 @@ def initialize_client(api_key=None):
     else:
         raise ValueError("Please provide an API key or set the TOGETHER_API_KEY environment variable")
 
-def encode_image(image_path):
-    with open(image_path, "rb") as image_file:
-        return base64.b64encode(image_file.read()).decode('utf-8')
+def encode_image(image_path, max_size=(800, 800), quality=85):
+    with Image.open(image_path) as img:
+        img.thumbnail(max_size)
+        buffered = io.BytesIO()
+        img.save(buffered, format="JPEG", quality=quality)
+        return base64.b64encode(buffered.getvalue()).decode('utf-8')
 
-def bot_streaming(message, history, max_new_tokens=250, api_key=None):
+def bot_streaming(message, history, max_new_tokens=250, api_key=None, max_history=5):
     if client is None:
         initialize_client(api_key)
 
@@ -31,7 +35,7 @@ def bot_streaming(message, history, max_new_tokens=250, api_key=None):
     messages = []
     images = []
 
-    for i, msg in enumerate(history):
+    for i, msg in enumerate(history[-max_history:]):
         if isinstance(msg[0], tuple):
             messages.append({"role": "user", "content": [{"type": "text", "text": history[i+1][0]}, {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{encode_image(msg[0][0])}"}}]})
             messages.append({"role": "assistant", "content": [{"type": "text", "text": history[i+1][1]}]})
@@ -50,19 +54,26 @@ def bot_streaming(message, history, max_new_tokens=250, api_key=None):
     else:
         messages.append({"role": "user", "content": [{"type": "text", "text": txt}]})
 
-    stream = client.chat.completions.create(
-        model="meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo",
-        messages=messages,
-        max_tokens=max_new_tokens,
-        stream=True,
-    )
+    try:
+        stream = client.chat.completions.create(
+            model="meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo",
+            messages=messages,
+            max_tokens=max_new_tokens,
+            stream=True,
+        )
+
+        buffer = ""
+        for chunk in stream:
+            if chunk.choices[0].delta.content is not None:
+                buffer += chunk.choices[0].delta.content
+                time.sleep(0.01)
+                yield buffer
 
-    buffer = ""
-    for chunk in stream:
-        if chunk.choices[0].delta.content is not None:
-            buffer += chunk.choices[0].delta.content
-            time.sleep(0.01)
-            yield buffer
+    except together.error.InvalidRequestError as e:
+        if "Request Entity Too Large" in str(e):
+            yield "The image is too large. Please try with a smaller image or compress the existing one."
+        else:
+            yield f"An error occurred: {str(e)}"
 
 demo = gr.ChatInterface(
     fn=bot_streaming,