That said, I'm planning future notebooks where you can perform text-to-image or image-to-image search, integrate OCR, scale it up, serve it, deploy it, etc.
96 karma · joined June 19, 2020
I also train AIs to write crappy Star Trek scripts and I build animatronic butterflies
That said, I'm planning future notebooks where you can perform text-to-image or image-to-image search, integrate OCR, scale it up, serve it, deploy it, etc.
At this rate we'll be releasing 4.0 next week
The trigram thing might just be a weird glitch in the pretrained model we're using (sentence-transformers/paraphrase-distilroberta-base-v1)
For image search we use Big Transfer Encoder (https://github.com/jina-ai/executors/tree/main/jinahub/encod...) but may switch to CLIPImage encoder at some point
Also swapped out the model
The framework is Jina (https://github.com/jina-ai/jina/) so it's pretty high-level. You can see the indexing/search Flow on lines 37-52 of https://github.com/alexcg1/jina-meme-search-example/blob/mai...
Had I known it would blow up I would've indexed more!
Really hoping Streamlit supports that in future.
This example is quick and dirty, indexing only about 1,000 images from a dataset we pulled from Kaggle. So if results suck, that's due to either the meme you search for not being in the dataset, or it just didn't get indexed in the random batch of 1,000.
I think the lack of syntax highlighting is going to be the major hindrance for me moving forwards.
Docs are never really "good enough" imho, but I like to think that ours are "good enough enough" for now (and only getting better )
A lot of people talk a big game about how awesome open-source software is. Not so many really understand its value and believe in it so much that they'd pack in their stable jobs at tech giants and risk it all for an open-source venture. We're in that second group.
A lot of people waste their words complaining about risk averse entrepreneurial and investment culture that hinders the rise of AI unicorns. Not so many want to shake up the system and fix it for the better. I think you can guess where we stand.
We are Jina AI, a venture-backed, well-funded AI COSS company, and we invite you to join us to build the next neural search ecosystem in open source. Will you step up?
Hope that helps. Let me know if you have any more questions or feedback!
PS: If you still have problems, could you let me know which browser you're using?