Closed sld closed 5 years ago
On default server config from https://raw.githubusercontent.com/deepmipt/DeepPavlov/master/deeppavlov/utils/settings/server_config.json I've got the same error.
Here config_path
should be a path to your model pipeline config, for example deeppavlov/configs/odqa/en_odqa_infer_wiki.json
If it's a config from our repository you can use just it's name without an extension , so it could look like
python -m deeppavlov alexa en_odqa_infer_wiki -d
If you want to override the default server config, you can define the DP_SETTINGS_PATH
environment variable. more info on this can be found in the settings documentation
Thank you! Finally, I've made Alexa work by using AWS Lambda Endpoint. I did https API calls from AWS Lambda to DeepPavlov models that ran through riseapi mode in my server.
Unfortunately, Alexa didn't work when I used https endpoint and run DeepPavlov in alexa mode. I've always got an error in Alexa side (The requested skill did not provide a valid response
) and no logs in DeepPavlov side (maybe some kind of SSL problem, but my certs are valid).
But it's not a big problem, because AWS Lambda worked.
I'm running in Alexa mode and getting the error:
My server_config.json:
Why is the error happening? Seems that everything should be right. I followed the documentation guide. Just deleted other skills from
model_defaults
section.