This channel aims at providing AI tutorials and hands-on learning experiences for everyone. In particular, I hope following messages come across clearly: 'Everybody can do AI' 'AI Made Easy' 'AI Innovation and Democratization' 'Grow your interest in learning AI and Data Science' 'Have fun with AI' !
I am posting the link to ppt/code as the description note under each video. Here it is, plus config file and some dummy sample files - feel free to use any documents you like for testing: drive.google.com/drive/folders/1bf6sKlfzfxm5QnVd0RA3MjIFpNK1j-jd?usp=drive_link
what could be more interesting friend if you can use local installations of models such gpt4all so all the data remain local !that would be fantastic !can you please give it a try ?
Certainly it is a brilliant idea to try gpt4all with all the benefits of running it on consumer cpu and keeping your data locally! In the meantime, will this open source have the same performance and generation quality as chatgpt? Reckon there is always a tradeoff there, depending case by case. But again it is a very good idea and I will find time to give it a try!
This question is not clear to me. If you meant tokens, yes, chatgpt (gpt in general) uses both tokenization and embedding techniques to convert natural language text/code into an abstract mathematical representation of data (called tokens e.g. sub-words and embedding vectors) ready for AI to learn and process.
Hello I'm new subscriber an I want to learn about Ai stuff, can you give me some advice? maybe what I have to learn , btw I'm do not have a basic programer skill Thanks you
Sure, here is some concise tips, hope it’s useful: 1 Pick up basic computer skill. Learn basic programming concepts and syntax, preferably using a data science language like Python. You don’t need to become an expert programmer, but it’s important to have a basic understanding of programming concepts. Check out the free learning resources I shared including Python, R, AI/ML/NLP ebooks and learning material here:drive.google.com/drive/folders/1ez7EKwF9doLA3cVFjZzpBS03xCXkvkYp 2 Familiarize yourself with the fundamentals of AI, including key concepts and terminology. Plus basic statistics. Google it and you will find many. 3 Take online courses and tutorials to deepen your understanding of AI. Some popular platforms include Udacity, Coursera, and edX. Search for data science, Python, machine learning, AI, NLP courses. Many courses are free to learn if you only ‘audit’ it. If you want, you can also deep dive to get certificate on some AI specialisation courses/mini programs. 4 Practice by building small AI projects. E.g. Following my tutorial series and hands-on notebooks step-by-step to do reproducible research and stay up-to-date on the latest developments in the field. 5 Join online communities to connect with other AI enthusiasts e.g. Kaggle
Yes, you can train a TTS system on your own data to produce more realistic voices. This is known as "speaker adaptation" or "speaker-specific training." By training a TTS system on speech data from a particular speaker, you can create a voice that is similar to that speaker's voice, producing more natural-sounding speech. However, training a TTS system on your own data can be challenging and requires a significant amount of speech data, as well as expertise in TTS and machine learning.
It is possible if you can in some way access Google Autosuggest corpus, which is supposed to be internal, huge and daily updated dataset of all Internet search queries.
Really it depends on the problem domain and text data. Try PKE first as it has a good zero-shot prediction performance if you don't know much about the text data to extract. Then check out the latest ChatGPT as per my introduction (ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-c0ano9D9Woo.html).
SEO is one of many typical use cases of keyword extraction. Yeah, surfer seo is a cool tool with 'keyword' built-in but it is a 'black-box', whereas many open-source tools were introduced here and you have full control on how/what keywords will be extracted for your specific purpose.
