Thank you, much better than most ! BUT i dont understand where to get the initial values of the matrices: 1) initial values for the matrix Mk used to create v1, ...., v4 v1, ...., v4 are the embeddings of the tokens in the sentence ...so Mk should be such that Mk x Identity matrix = v1, v2, v3, v4 ...this would make math sense, but is not what you (or anybody else) say 2) initial values for the matrix Mq used to create the queries v1, ...., v4 initially the queries are the same v1, ...., v4 vectors ..so Mq should be the same Mk so that Mq x Identity matrix = v1, v2, v3, v4 ...this would make math sense, but is not what you (or anybody else) say 3) initial values for the matrix Mv used to create the "values" y1, ...., y4 ...that could be the Mv containing the normalized dot products Wij ...but again i doubt that this is what you mean to say

Can I connect it to twilio voice? for phone interaction?

you deserve more fans!~

Bravo, bro!~

i already receive the email but it has no license key

Another heads up, this was the first recommended tutorial video, but at the end of this video you link to the Last video in the series. Doesn't seem like the obvious next step. :)

this tutorial is not for those without coding knowledge

from 4:47 you did not mention what note you are writing from or coding from aside the system default terminal machine .

What a great explanation about this topic. Many thanks

How do I install rasa pro using license key 5:09

awesome :3

is there a github repo which has all this?

Thank you so much, you made me happy.. even after 3 years feels relevant to me

I think in [5:14]: it is endpoints.yml and in [11:02] we have python file and not domain.yml.

Yeah its very odd that this doesn't have more views. It's almost like people are so addicted to hype, they don't want to bother understanding these simple, elegant mechanisms, and instead have baloney conversations about AGI, etc. etc. 😕 Based on my understanding gained from watching this series, I have developed some ideas about how to improve LLMs and been awarded an NSF ACCESS (supercomputing time) allocation, and submitted an NSF SBIR proposal improving these models. Great job!

This could have so many use cases, thanks for making s tutorial

Thank you for the update! Very Helpful!

Which app is used to create this presentation?

The link doesnt work anymore i guess

If the user gives the time of da😂tes along with the first input, is calm capable of capturing it? I would like to book vacation time for Jan 10 to 20 5:55

i have done everything correctly but when i execute 'rasa -h' ti chows that there is no module named 'randomname' , even if i install 'randomname' separately it doesnt work ... what should i do ?

There are a lot configurations on the toml file that are not by default. It would be great to mention that. Besides, it seems that the template now is deprecated

Nice shelves.

Intersting video. Does this code work for Rasa 3.6.20 too?

great series

Very intersting! I'm building a booking chatbot base prompt and calling function use OpenAI. But, many issues like unstable token generation speed, halu...Does Sara can resolve them in a better way?

Thank you so much! This is an excellent playlist for getting into the topic of CALM. Justina is wonderful; I could listen to her endlessly :) Special thanks for making this solution open source :)

This is great explanation of KQV in attention and transformer logic.

0:35 In credentials.yml launches an error: RasaException: Failed to find input channel class for 'alexa_connector.AlexaConnector'. Unknown input channel. Check your credentials configuration to make sure the mentioned channel is not misspelled. If you are creating your own channel, make sure it is a proper name of a class in a module. Rasa version: 3.6.15 Python version: 3.10.13

Around 6:05 you mentioned (by introducing weights) "maybe we can learn more patterns". You did not mention why or how using weights are helpful. One of the reasons is that without introducing weights and we stick to the dot product, using your example, the product between v_x and v_x will always be the highest as they are the same token, and the dot product between two tokens are always pre-determined by your pre-trained word embeddings and are fixed (e.g. the embedding for 'bank' is fixed and will not change). By introducing weights, you are allowing the network to learn/update the inter-token relations by seeing training examples.

I like the sound of this.

Can we integrate chatGPT API with RASA framework ?

@rasa Did you mean endpoints.yml ? @5:13

i have issue when python -m pip install -U pip i had warning what should i do?

Awesome

I've been looking all day for a clear explanation of the K, V, and Q concepts. Finally found it. Nicely done!

where do you place your restaurant.csv in your project?

Great content! 😊

90% of those who talk about this field do not actually know the subjects, they just tell what they have seen from somewhere. It is not difficult to understand this if you are an experienced learner. For the first time, I came across someone who explained the subject with real knowledge. Thank you endlessly.

hey there, i have a problem, when i change OPENAI api to GEMINI api it still asks for openai api and throws error in your template tutorial

Hello, how do you run rasa calm in production ? Should we install rasa-plus or rasa-pro ?

What a beautiful video!

Thank you for sharing the video with us. I really appreciate it

Thanks for these guidance.

very clear thanks

You can use the Thunder client vs code extension alternative for postman

This is amazing.

How do I keep it running with my app or do i have to sperately turn the rasa bot on each time in the moodbot directory?

its off topic but me, as a human, thought bank of the river as a bank with a cool name lol

Hi. This is a really great visualisation of weightings - thank you! I have a question though: at 11:30 you say you're going to calculate the cross product between the first token's vector and all the other vectors. Should this instead be the dot product, given that you are looking for similarity?