NLP engineers // NODE-RED enthusiasts // co-founder // Germany
Transformers, TensorFlow, NODE-Red, n8n. All these buzzwords are nice.
But do you wanna build for the masses? Do you prototype and see the biggest smile on a customer face?
If you do, hit me up then I'll share some insights. Cheers
I am hugely interested in working on a NLP based project. I am already working on one in my company and would like to collaborate. I am based in the U.S. Please let me know how does it work out.
Oh love it. NLP is and will become huge. I am just follwing the latest research, cause of my thesis. And it is so fast. very big player is involed.
Cheers.
I work on delivering a NLTK, Spacy like library which involves sentiment, tokenization, entity, keywords etc. features. I have trained sentiment models before but not entity. Will be super interested to do so.
Hey there, how did you get immersed in NLP? Is it accessible for junior developers in your opinion?
Thanks!
Hi @PseudoWalrus, as NLP is firstly a broad area and secondly evolving so fast, your question is super relevant.
I am still facing the same problem. And it seems, that reading papers is working better than try to read code.
So a few sources I found helpful:
Of course there are thousand book, but I would start by getting an overview, learning about vocabs, benchmarks, methods and tasks before actually trying to code.
You can use the Feynman technique (https://doist.com/blog/feynman-technique/).
How was your approach so far, when learning new stuff?
Hey mate,
Thanks for the resources, got lots to sink my teeth into now!
I feel it's hard to balance learning the theory and putting it into practice, in terms of allocating time. The code is quite intimidating at the moment so I think I will stick to reading for now.
My approach is generally to skim through a book, extract the core concepts, then attempt to visualise and apply them if possible. Afterwards I try to identify the gaps in my understanding and target those in more detail. Doesn't always work out tho, lol.