When I first started using Google Bard (now renamed Gemini), I hoped to be able to upload my local pdf files for conversation, but I did not find the entrance to upload files and could only upload images. This doesn't meet my needs.
So I started looking for a solution, if I could upload my pdf files in Bard and be able to have a conversation.
After some searching, I found someone saying that you can use an online pdf link, enter it into the input box, and then ask a question. I tried it using someone else's link and it seems to work, which is great. So I uploaded the file to my website, then sent the link to bard, and started trying to have a conversation based on the file content.
It looked good at first, but later I discovered that this function was not really implemented.
Just an illusion. If you use an online PDF file in the prompt word, Bard or Gemini will only answer based on the name of your online PDF document, and then based on its knowledge base or online search for relevant content, not based on the actual content of the document. For example, the name of my PDF document is 'How to use PDF for conversation in bard', but the content of the document is 'How to make a pizza in 7 steps'. When I ask him to summarize the content of the document or answer based on the content of the document, The content of his answer is always 'how to use pdf for conversation in bard'.
So I started a new attempt and finally found the correct way to use it: Extensions based on Google Workspace of Bard or Gemini to talk to PDF files.
Therefore, through this project, I want to introduce to you how to use pdf files for conversations in Bard or Gemini. Share my experience with everyone. Don't be misled by wrong usage methods again.
So I started looking for a solution, if I could upload my pdf files in Bard and be able to have a conversation.
After some searching, I found someone saying that you can use an online pdf link, enter it into the input box, and then ask a question. I tried it using someone else's link and it seems to work, which is great. So I uploaded the file to my website, then sent the link to bard, and started trying to have a conversation based on the file content.
It looked good at first, but later I discovered that this function was not really implemented.
Just an illusion. If you use an online PDF file in the prompt word, Bard or Gemini will only answer based on the name of your online PDF document, and then based on its knowledge base or online search for relevant content, not based on the actual content of the document. For example, the name of my PDF document is 'How to use PDF for conversation in bard', but the content of the document is 'How to make a pizza in 7 steps'. When I ask him to summarize the content of the document or answer based on the content of the document, The content of his answer is always 'how to use pdf for conversation in bard'.
So I started a new attempt and finally found the correct way to use it: Extensions based on Google Workspace of Bard or Gemini to talk to PDF files.
Therefore, through this project, I want to introduce to you how to use pdf files for conversations in Bard or Gemini. Share my experience with everyone. Don't be misled by wrong usage methods again.
Hope you enjoy it.