Hi, first of all thanks for making this summarizer public.
I was playing around with different models for extracting summary on case data ( customer messages - contains lot of rants )
The summary in the demo https://smrzr.io/ was better than what I could achieve by using roberta-base, distillbert, bert, bart-large model on your summarizer.
Can you please tell what model is used in the demo url ? I see the ratio used is 0.15
P.S. I saw the difference on this paragraph :
I want to talk to agent. I am facing internet issues. This is the worst internet service. Your internet sucks.
The models which I use generate no summary, but the demo url generates I am facing internet issues
Hi, first of all thanks for making this summarizer public. I was playing around with different models for extracting summary on case data ( customer messages - contains lot of rants ) The summary in the demo https://smrzr.io/ was better than what I could achieve by using roberta-base, distillbert, bert, bart-large model on your summarizer.
Can you please tell what model is used in the demo url ? I see the ratio used is 0.15
P.S. I saw the difference on this paragraph : I want to talk to agent. I am facing internet issues. This is the worst internet service. Your internet sucks.
The models which I use generate no summary, but the demo url generates
I am facing internet issues