Vladimir Blagojevic

@shadeMe I think this is the solution we had in mind. LMK

Really cool to have these options 🚀

@danielbichuetti thanks for raising this issue. We intend to support Bing via https://serpapi.com/ Does SerpAPI satisfy this feature request?

Ok cool, lets see if we can create some common components for both, perhaps even reuse them further in other third party APIs. Let me get back to you after...

@sjrl should be IIRC, let us know otherwise.

Would you please share your example @ss2342 ?

@ss2342 IIRC the necessary callbacks to capture stop words are not called from quantized models. I'll double check once again. But regardless, we should do a better job in that...

It should work with many other generators that support stop words, can you use them? Or can you somehow not use quantization? :-)

Ok thanks @ss2342 I'll work on this next week until it is solved.

This works for me repeatedly, verbatim example from above. Here is the notebook https://github.com/vblagoje/notebooks/blob/main/hf_stop_words_test.ipynb Please advise @masci