Hi Karndeep. Thank you for this. Question, does Gemini give you a confidence level or anything related? How can you be certain about the extraction result since i think the model is an LLM. Do other alternatives such as DONUT or Layoutlm give you a confidence level?
Namaste sir I am a blind person. Want to learn Data annotations and labelling from zero How to get started?Can I get freelance opportunities in this domain with the required skills?
The Gemini 1.0 Pro Vision model will be deprecated from Google AI services and tools as of June 12, 2024. Usage of the model in Vertex AI is not affected by this notice. What this means for you You’ll be able to use Gemini 1.0 Pro Vision until July 12, 2024. After July 12, saved prompts using Gemini 1.0 Pro Vision in Google AI Studio will switch to using Gemini 1.5 Flash. API calls that specify Gemini 1.0 Pro Vision will fail.
I cant run my model in office environment in Jupiter notebook it showing tcp is shutting down i need suggestions for running it in local jupyter notebook
If there are some complicate layouts then it is advisable to finetune the models like LayoutLM and DONUT. Also, LayoutLM is OCR dependent so it may struggle to extract some information from complex documents where in DONUT is OCR independent and it can be much better choice in such cases
Hii karandeep , big thanks for this meaningful content i am a student and i want to make a project is extract text from a hindi admission so how can I do ? use any model or some kind o fine tune model or vice versa ?
sir plzz respond as it is possible . using this model and ai iam getting result with the 90% accuracy iam not getting 100% accuracy some of the inputs are scanned copies also there either it is scanned copy that is not matter but 100% accuracy iam not getting i changed the params also any suggestion sir plzz
I had completed my company task successfully with the help of these videos
Same Here
@@rohanborse5268 how to extract key and value pair from markdown object
Hi Karndeep. Thank you for this.
Question, does Gemini give you a confidence level or anything related? How can you be certain about the extraction result since i think the model is an LLM.
Do other alternatives such as DONUT or Layoutlm give you a confidence level?
thanks
Thank you so much sir ❤❤❤❤
Namaste sir
I am a blind person. Want to learn Data annotations and labelling from zero How to get started?Can I get freelance opportunities in this domain with the required skills?
The Gemini 1.0 Pro Vision model will be deprecated from Google AI services and tools as of June 12, 2024. Usage of the model in Vertex AI is not affected by this notice.
What this means for you
You’ll be able to use Gemini 1.0 Pro Vision until July 12, 2024.
After July 12, saved prompts using Gemini 1.0 Pro Vision in Google AI Studio will switch to using Gemini 1.5 Flash. API calls that specify Gemini 1.0 Pro Vision will fail.
I cant run my model in office environment in Jupiter notebook it showing tcp is shutting down i need suggestions for running it in local jupyter notebook
Great video Karndeep. My understanding is that Gemini is better in Document AI data extraction than GPT4. Any comments on this?
I have also seen some amazing results with Gemini. GPT4 seems to be good for general and simple layout documents.
Hi Karndeep. Great content as always.
How do you think Gemini aí compares with layoutlm or donut (NN for token classification with layout info)?
If there are some complicate layouts then it is advisable to finetune the models like LayoutLM and DONUT. Also, LayoutLM is OCR dependent so it may struggle to extract some information from complex documents where in DONUT is OCR independent and it can be much better choice in such cases
Hii karandeep , big thanks for this meaningful content
i am a student and i want to make a project is extract text from a hindi admission so how can I do ? use any model or some kind o fine tune model or vice versa ?
can you tell me how to pass more than 1 images as input and than get output based on the prompt
sir plzz respond as it is possible . using this model and ai iam getting result with the 90% accuracy iam not getting 100% accuracy some of the inputs are scanned copies also there either it is scanned copy that is not matter but 100% accuracy iam not getting i changed the params also any suggestion sir plzz
Hi, could you do make an app which is classify pdfs with machine learning or ocr but pdfs are scanned
this api is for free from google or limited?