Rasa: Rasa API / CLI community feedback

Created on 7 Mar 2019  路  7Comments  路  Source: RasaHQ/rasa

Hi there 馃帀

If you tried out the new Rasa API / CLI, please leave any feedback (bugs, suggestions) in this issue.

Thanks for your support!

stale type

Most helpful comment

Awesome! This step with the Rasa API / CLI will make the chatbot development with Rasa much easier to beginners. It's clear that many changes have been done and it's clear too, the care with each piece of code, congratulations!

All 7 comments

I'd probably add to the instructions the following:

python3 -m venv .venv
source .venv/bin/activate
pip3 install -r requirements.txt

Or something like this just to rule out any issues you might see from people with newer tensor flow libraries, etc.

Everything went great with the test I like this a lot and the CLI is gonna make stuff super easy to get going without having to build the structure out.

As a learning thing I'd probably have a bot ready to go with a basic conversational setup so you can also offer people an option in the CLI to load a base model that already has the basic pleasantries done maybe like Hello, Goodbye, etc.

Overall solid work and I'll be using this once it goes live for sure.

Hello, guys, I tested the rasa stack and added my jupyter-notebook in my GitHub, you can read it here I made some comments about it and in my jupyter has my reasoning I think it's a better way to understand how other people will use it.

summing up, I really liked the rasa init and the tests are important but it needs time to understand all the test output.

If have any other activity to do about it, please let me know.

@arthurTemporim

I did not like the way that the tests outputs are given, it's hard to understand and find issues, maybe some green and red colors can make this easier.

Awesome idea! We should definitely do that

what means "Speak to separately trained models"?

The new commands put the core and nlu model in one tar-file, but we still want to be compatible with the old way (training core and nlu separately)

I didn't find documentation about the rasa_core.test what is the difference between this command and Rasa e2e tests?

We will create some documentation on it as soon as we finalize the commands. If you set the parameter e2e=True it's the same command as the end to end tests. We just wanted to make the API easier to use that's why it's rasa.test(...) and not rasa.test.test(...)

Awesome! This step with the Rasa API / CLI will make the chatbot development with Rasa much easier to beginners. It's clear that many changes have been done and it's clear too, the care with each piece of code, congratulations!

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

We can close that, the goal of this issue was to iterate on the CLI before we publish Rasa 1.0, what's what happened

Was this page helpful?
0 / 5 - 0 ratings

Related issues

yondu22 picture yondu22  路  3Comments

nikhilcss97 picture nikhilcss97  路  3Comments

jahid-ict picture jahid-ict  路  3Comments

karnigili picture karnigili  路  3Comments

nahidalam picture nahidalam  路  3Comments