Recently we started to observer problem with Dialogflow: We have an agent with ~200 intents. Everything works as expected, but then, all of the sudden, without changing anything in the agent, intent that was matching with some input, just stops matching with the exact same input. The only way to fix it is to go and manually click train and then it works again.
Any thoughts or does anyone had the same issue?
Related
I have some issues from my Agent. It often happens that what the “user says” fall into the wrong intent. So I am trying to training the agent by the specific section “TRAINING” (please see the screenshot). I am changing the wrong intent to fix the issue and when I click on the APPROVE button then nothing happens! The problem is that training does not start (the wheel gear does not turn). And the wrong intent returns again. So the Agent continues to choose wrong intent in some cases. I would know why this training section not is working …
I'm working on a pre-existing python code-by-zapier zap. The trigger is "Code By Zapier; Run Python". I've made some changes to the contained python script, and now when I go to test that step I run into the following error message:
We couldn’t find a run python
Create a new run python in your Code by Zapier account and test your trigger again.
Is there any way of figuring out what went wrong?
I'm guessing a little bit, but I think this issue stems from repeatedly testing an existing trigger without returning a new ID.
When you run a test (or click the "load more" button), then Zapier runs the trigger and looks through the array for any new items it hasn't seen before. It bases "newness" on whether it recognizes the id field in each returned object.
So if you're testing code that changed, but is returning objects with previously seen ids, then the editor will error saying that it can't find any new objects (the can't find new run pythons is a quirk of the way that text is generated; think of it as "can't find objects that we haven't seen before).
The best way to fix this depends on if you're returning an id and if you need it for something.
Your code can return a random id. This means every returned item will trigger a Zap every time, which may or may not be intended behavior.
You can probably copy your code, change the trigger app (to basically anything else), run a successful test (which will overwrite your old test data), and then change it back to Code by Zapier and paste your code. Then you should get a "fresh" test. Due to changes in the way sample data is stored, I'm not positive this works now
Duplicate the zap from the "My Zaps" page. The new one won't have any existing sample data, so you should be able to test normally.
I created a very simple agent on dialogflow. Every now and then, even using static responses (where only the session ID changes), the application crashes (both on Google Assistant and in the simulator).
On the logs I find the message in question: "MalformedResponse at expected_inputs [0]: Responses must provide at least one possible if this is not the final response. What does it mean?".
I don't understand what it means.
Could anyone help me?
I'm having the same problem. That i've noted is that this problem doesn't ocurre every time. The behavior is really unstable.
I'm using 2 required parameters in the intent. I don't know if that has something to do with the problem.
MalformedResponse at expected_inputs[0].input_prompt: 'input_prompt' is empty
MalformedResponse at expected_inputs[0]: Responses must provide at least one possible intent if this is not the final response
The problem, appears to be caused when you use own parameters, not when use sys. parameters.
Google answer me the following:
"Hi Alberto,
This issue has already been escalated to our engineering team. You are correct about required parameters causing the problem. Our engineering team is working thoroughly to resolve this since it is affecting some other AoG projects as well. "
I was facing the exact same issue when I was using custom entities as input parameters to my action.
I was asking for 2 custom parameters and one system (sys.color) and triggering the same with input prompts.
I found a workaround for that by taking all 3 in separate intents and using the response of the previous intent as a trigger to enter the next parameter, also by making the training phrases as sample values of the custom parameter.
I am still in contact with Dialogflow Suppport and waiting for bug fix but it works for now.
During my process there is a drop-down html element that I have spied and set correctly. I run through the Object and Process studio without any issues. Once the same process is ran through the Control Room the element throws and error that it cannot be found.
I have tried multiple different configurations without any luck and the element is still found without issue when ran manually. I even checked by signing into the VM having it error and on the second retry had the VM up; at this stage the element was found without issue. It seems to only be when the bot is running and the screen is not up. No other elements give this issue and the next step is the same dropdown, but for a stop time not start.
Any help would be appreciated!
I have added pics of the STARTSPIED - START and ENDSPIED - END spied configs,the Navigation Stage, and the process correctly running in the Object Studio OB Studio correct.
Thank you!
For the person who down-voted this item...Configs here are my two days of config changes and research that I tried on my own before asking for help. This is my 5th automation that I have put into deployment without help, so thank you for down-voting someone trying to get help where they are stuck.
Your problem is most likely connected with the fact that processes run in the control room are executed much faster than in studio and your webpage might not be loaded on time (that would explain why it works after retry).
Best practice approach would be to add a dynamic wait stage after attach and use “Parent Document Loaded” option on the element you want to interact with. It will wait for the page to be loaded and then check exist for the element. I would also suggest splitting your action into two, first to set start date and second for end date.
I am using my ccnet which is configured with the clear case
and everything is working fine as expected,but i am unable to see
anything in the "ViewProjectModificationHistory".whenever i click this
tab,i see a message which tells that :"No history Data found, make
sure you use the modificationHistory Publisher for this
project".Infact i had added under the tag.And to my
suprise,when i open this same link in the other persons system,it
opens well and fine and see all the modification history recorded.
So can anyone please tell me what could be this issue and how can it
be resolved?
Thanks and regards
Maddy
You should check your build history file. This will contain the XML data that feeds these reports. In particular, check the modifications/modification nodes to see if they are populated or not.
My guess is that the modifications are not there - so there may be nothing to display.
Another gotcha I have caught myself doing is not actually checking that there was in fact changes committed since the last time I built a project.