Merge pull request #8 from forensicmike/dev

Readme update and small fix to image loading
This commit is contained in:
MagicSource 2025-02-25 10:25:11 +08:00 committed by GitHub
commit 7a5ff97458
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 8 additions and 1 deletions

View File

@ -3,6 +3,7 @@ from PIL import Image
import base64
from threading import Thread
import io
import os
from transformers import TextStreamer
try:
@ -12,6 +13,8 @@ except ImportError:
def load_image(image_file):
if isinstance(image_file, Image.Image):
return image_file
if image_file.startswith("http") or image_file.startswith("https"):
response = requests.get(image_file)
image = Image.open(io.BytesIO(response.content)).convert("RGB")
@ -21,7 +24,9 @@ def load_image(image_file):
def load_multi_images_maybe(image_files, splitter=" "):
if isinstance(image_files, str):
if isinstance(image_files, Image.Image):
return [image_files]
if isinstance(image_files, str) and not os.path.exists(image_files):
images = image_files.split(splitter)
else:
images = image_files

View File

@ -68,6 +68,7 @@ pip install -U namo
A simple demo would be:
```python
import torch
from namo.api.vl import VLInfer
# model will download automatically
@ -76,6 +77,7 @@ model = VLInfer(
)
# default will have streaming
# this method accepts a single path, url, or PIL image, or an array of same.
model.generate(images='images/cats.jpg', prompt='what is this?')
```