@Sunil_Natraj In one line - it's rocking 
Thanks a lot for the 'History' tab.
One more option will be helpful I think - Throttle delay (in milliseconds) - as those who are using free tier APIs (like me 
) has often a rate limit. Such an option may improve our situation.
-partha
Open refine 3.9 worked, but no LLM provider list is available. How can I configure LLM providers?
Congrats for the great work.
B.
Thank you @psm. Glad this makes the extension more useful. Your input on support for supporting delay between requests makes sense, can you add this request here
Hi, There are no pre-defined LLM Providers, you will need to define them, refer the documentation.
@belfra Could you plz save it (configuration) first and then test the Ollama?
It should then provide response like this:
-partha
Must be something in my LLM url: http://localhost:11434
Something is missing...
Great, great work. congrats u all
B.
@belfra May try this (what I'm using for Ollama):
[in place of the model llama3.1:8b use your own model - ollama list command will provide the model names that you are using. Use the exact name under the Name column]
Thank you so much, you are the MAN!
B.
AI Extension V0.1.2 is now available for download. Link
Thank you for all your inputs and support in this release.
@Sunil_Natraj @psm it may be good to update the AI - LLM Provider Definition Guide page with ready to use template for the different LLM provider.
@Sunil_Natraj and @Martin, good idea! If I provide the content in a word processor (whatever services I've attempted so far), will that suffice?
Yes that will be helpful. I can add that into the help guide.
@Sunil_Natraj and @Martin
Is this model okay to follow for other LLM providers?
Plz let me know.
NB: Upload is not allowing .odt/.docx (only pdf format)
llm-services.pdf (65.3 KB)
Hi @psm This looks good. PDF format is fine; I will be able to create the MD format from it
Please find the first version (to be updated gradually, lets start with this).
Regards
Thanks @psm I have published guides for Ollama, Groq and OpenRouter and links included in the main help page
As someone quite familiar with OR and conventional data wrangling, but not in Python I also want to thank you for this very exciting lesson, which finally got me to take a closer look at Python. With deepseek-r1-distill-llama-8b I've achieved good results for my purpose in the first tests (I tried different LLMs but my challenges were not model related in the end).
Learnings:
Learned a lot, thanks!
Hi @colognella, thanks for the feedback! Have you also tried the great LLM Extension by @Sunil_Natraj? It makes life much easier when dealing with local (but also remote) models.
