ai-workspace-services/litellm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

added decoding of base64 image data in _gemini_vision_convert_messages function

Browse Source
This commit is contained in:
hugo.lima 2024-05-17 16:46:56 +01:00
parent 186f11612a
commit 32e25cdb11
1 changed files with 17 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
litellm/llms/prompt_templates/factory.py
View File

@ -1328,6 +1328,7 @@ def _gemini_vision_convert_messages(messages: list):
# Case 1: Image from URL
image = _load_image_from_url(img)
processed_images.append(image)
else:
try:
from PIL import Image
@ -1335,8 +1336,22 @@ def _gemini_vision_convert_messages(messages: list):
raise Exception(
"gemini image conversion failed please run `pip install Pillow`"
)
# Case 2: Image filepath (e.g. temp.jpeg) given
image = Image.open(img)
if "base64" in img:
# Case 2: Base64 image data
import base64
import io
# Extract the base64 image data
base64_data = img.split("base64,")[1]
# Decode the base64 image data
image_data = base64.b64decode(base64_data)
# Load the image from the decoded data
image = Image.open(io.BytesIO(image_data))
else:
# Case 3: Image filepath (e.g. temp.jpeg) given
image = Image.open(img)
processed_images.append(image)
content = [prompt] + processed_images
return content