You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
hi @antoan this should be possible as of #834 or v2.2.2 - does that seem like what you need?
example:
In [1]: importmarvinIn [2]: link= "https://plus.unsplash.com/premium_photo-1674917000586-b7564f21540e?q=80&w=4912&auto=format&fit=crop&ixlib=rb-
...: 4.0.3&ixid=M3wxMjA3fDB8MHxwaG90by1wYWdlfHx8fGVufDB8fHx8fA%3D%3D"
In [3]: importhttpx
...:
...: marvin.beta.caption(marvin.beta.Image(data=httpx.get(link).content))
Out[3]: 'A serene sunrise over a mountainous landscape with a winding road and alpine wildflowers in the foreground. The mountain peak is partly veiled by light morning mist.'
here im getting bytes from the image (which happens to live at an external url, just because that's easier to share), which you could swap for your own image's bytes
Is it currently possible to pass encoded image strings to Marvin, to extract structured data, rather than external urls?
Any thoughts would be much appreciated.
The text was updated successfully, but these errors were encountered: