No problem! It will be out tomorrow. I had planned to release it last week, but I had a few audio/video issues when I got the recordings into post-production. Thanks for watching the videos and I am pleased to hear that you are enjoying the series!
Nice videos, learning a lot and I am very interested in any work about books/literature. Do you know if Prodigy can be fully utilized (GPU) on a MacBook Pro M3 Max 128GB memory? I think is spaCy is using my GPU's(mps), it's hard to tell. Seems like Apple Silicon is either too new or just forgotten in some of this newer software. Thanks.
Thanks! I am coming back to this series shortly. I am not sure. I have a similar device, just 64 gb version. I will do some tests. I expect not, but I expect it to follow so in the near future wil mlx
I have been eagerly awaiting this video. When will the next one be released? I have been using Prodigy in my master's thesis ever since they introduced a license for researchers. And this series has been very helpful to me.
Thanks so much! Much sooner. I was preparing the next few videos at the same time while trying to finish a few projects at work. That's why it took so long for this video. The next one will be out early next week. Thanks for your patience and I am so happy to hear that you are finding this series useful!
Would love more prodigy based content, especially anything using nested spans -- also did you ever push the rest of the project to the repo or just step 1?
Do you know how to manually cultivate data? I've been watching almost all of your videos, but none of them show how to train or explain how to cultivate the DocBin for SpanCAT and utilize the command to train the spaCy model without purchasing or using any annotation data framework. Each time I try to create the example and then save it in the DocBin, it gives me the following error: [E855] Invalid span: span is not from the same doc. But I can't seem to find the fault. It's like searching for a needle in a haystack! By the way, thanks for your contribution! appreciate it.