Thank you for contacting me. Please note that I live in the GMT+3 time zone - responses might be delayed by this.
-
AuthorPosts
-
-
January 9, 2024 at 11:45 pm #9567
Hi,
See this video:
I am getting this error:
[9-Jan-2024 21:40:53 UTC] Error: Invalid request submitted to the completions API, result: stdClass Object ( [message] => The model
text-davinci-003
has been deprecated, learn more here: https://platform.openai.com/docs/deprecations [type] => invalid_request_error [param] => [code] => model_not_found )1. Article is generated however meta description is enabled but not generated.
2. I am not using
text-davinci-003
anymore, the plugin is upto date but it s a old campagone -
January 9, 2024 at 11:48 pm #9568
This reply has been marked as private. -
January 10, 2024 at 3:11 am #9569
Hi,
Additional info, I saw this in log tab. This was in Omniblock
8-Jan-2024 21:56:53 UTC] Starting editing post ID: 77073
[8-Jan-2024 21:56:53 UTC] Calling OpenAI (\text-davinci-003) Post Editor with seed command: Write a formal article in English about: “Mastering the Art of Feline Obedience: The Ultimate Guide to Cat Training Tools
”
[8-Jan-2024 21:56:53 UTC] Error: Invalid request submitted to the completions API, result: stdClass Object ( [message] => The modeltext-davinci-003
has been deprecated, learn more here: https://platform.openai.com/docs/deprecations [type] => invalid_request_error [param] => [code] => model_not_found ) -
January 10, 2024 at 6:16 am #9570
Hello,
First of all, thank you for your purchase.
I updated the plugin on your site, please check if the issues are solved. Let me know.
Regards, Szabi – CodeRevolution.
-
January 11, 2024 at 1:07 am #9589
I have been get this error now:
[10-Jan-2024 06:13:03 UTC] Failed to list OpenRouter models: OpenRouter API key not set up in plugin settings.
-
January 11, 2024 at 3:42 am #9590
This reply has been marked as private. -
January 11, 2024 at 9:41 am #9591
Hello,
0. The latest plugin update will fix the error you mentioned.
1. Please check the Activity and Logging menu of the plugin -> ‘Activity Logs’ tab. You should see there more info on why no meta description was generated in some cases.
2. This is currently not possible, but I noted this idea.
3. This is currently not supported, but I noted this idea, will think on it for upcoming updates.
4. You need to specify in the prompt which is sent tot he AI writer, which generates the conclusion, that you want only a single section generated.
5. Be sure to UNCheck the ‘Don’t Try To Fool AI Detectors (Disable Content Tricks)’ checkbox from ‘Main Settings’ -> ‘General Settings’ tab.
Regards.
-
January 11, 2024 at 12:36 pm #9594
For item 5, If have it unchecked, then it will not be detected as AI content. Right?
-
January 11, 2024 at 12:48 pm #9595
This reply has been marked as private. -
January 11, 2024 at 2:25 pm #9598
I enabled “Detailed Logging” in the plugin for you, more logs will appear now. Also, when the circle is running for a rule, don’t run it again, as this means that it is already running. Trying to run an already running rule will return the yellow status, which is normal.
Please check logs after the plugin starts execution.
Regards.
-
January 11, 2024 at 7:30 pm #9599
Thanks.
For item 5, If have it unchecked, then it will not be detected as AI content. Right?
-
January 11, 2024 at 8:05 pm #9600
Correct.
-
January 12, 2024 at 11:01 pm #9604
This reply has been marked as private. -
January 13, 2024 at 6:46 am #9607
Hello,
Well, the only method to improve content quality in these cases is to do experiments with different variations of prompts which are sent to the AI writer.
I suggest you change the prompts in multiple ways (maybe use also the help of ChatGPT, asking it how to improve the AI prompts that you currently have set up in the plugin, to not generate the above content issues any more).
After experimentation, I recommend you keep the prompts which work best for your use case.
I hope this helps.
Regards.
-
January 13, 2024 at 12:42 pm #9611
This reply has been marked as private. -
January 13, 2024 at 12:55 pm #9612
Yes, there are some variations in the generated content, even with the same prompt.
I suggest you try experimenting also with different AI models, as other models might provide better results.
You can try using even Anthropic Claude, Google Gemini Pro or OpenRouter models for this.
-
January 15, 2024 at 4:00 pm #9672
This reply has been marked as private. -
January 15, 2024 at 4:45 pm #9673
Thank you for pointing this out, I noted it, will think on a way to remove the quotes and improve headings.
-
January 16, 2024 at 7:36 pm #9694
Hi.
Is the above issue fixed in the latest update? It seems, the issue still exists.
Thanks
-
January 17, 2024 at 5:23 am #9695
It still needs more work, as the AI returns these characters in an unpredictable way. I need to do more prompt experiments.
-
January 18, 2024 at 3:02 am #9696
Please fix this, as I have been stuck and I can’t generate articles.
Thanks
-
January 18, 2024 at 7:14 am #9698
Please send me temporary admin login credentials to your site and I check on this. My email is kisded@yahoo.com
Regards.
-
-
AuthorPosts
The topic ‘I get a few errors’ is closed to new replies.