Installation issue in windows10,python 3.7 and ms v c++ 2017

Hi All, I’m facing issue while try to install rasa nlu. Have windows 10 pro, python 3.7(64), pip 18.1 and ms vc++ 2017 (64) 14.16.27012. Getting fatal error c1083: cannot open include file: ‘io.h’: no such file or directory error:command ‘visual studio path’/cl.exe failed with exit status 2

Hey did you try this! Actually compiler is looking for a directory that is not the pwd. you have to set path in vc++. check out this post

hey @thangavel i have created a video which will help you solve the issue :slight_smile:


Hey @JiteshGaikwad How is it going? :slight_smile: I have a query. can you take a look please.?

hey @Akshit i think i had answered your question already in this post :smiley:

Isn’t that method inefficient!? already i’m passing sender_id in the GET request! Can I not just take it from there just like I’m taking query! How can i do that?

Hi ,

Thank you :slight_smile: its worked. May I know Anaconda python is recommended or shall we use traditional python?

hey @thangavel Unless you’ve already got numpy & scipy installed, Rasa highly recommends that you install and use Anaconda. but i prefer using traditional python :sweat_smile:

Thank you :slight_smile:

Hi , I have following error while using tensorflow_embedding. But I have installed tensorflow. please review.


hey @thangavel i had also faced the issue for the first time when i had setup the rasa on my windows system, so as a solution for this, I just uninstalled python and reinstalled it again & it worked well and after that i din’t faced the above issue :sweat_smile:


I’m getting same exception after re-installed python 3.6.

did u uninstalled the tensorflow?

yes I did

I also struggled with this,

I would install anaconda or python 3.6. Then install the Visual studio windows 10 sdk and the C++ components. I also had to install pywin32

and then the spacy commands for the english language pack

pip install -U spacy python -m spacy download en

Restart at all the prompts.

then start with setting up a virtualenv project

Then either choose core or nlu or both

and follow the commands on the instructions pages such as an example : git clone cd rasa_nlu pip install -r requirements.txt pip install -e .

Once you finished you run the trainng commands and this should be enough to get you going.