Google Gemini Function Calling 🚀 EASILY Integrate your OWN Function!

Key Takeaways

this is amazing what if you can do function calling in Gemini large language model what if you can integrate that in your python application that's exactly what we're going to see today let's get [Music] started hi everyone I'm really excited to show you about function calling in Gemini large language model I'm going to take you through step by step on how to do this but before that I regularly create videos in regards to Artificial Intelligence on my YouTube channel so do subscribe and click the Bell icon to stay tuned coming to using Gemini API using vertex AI we've already covered this in a previous video which I will link that in the description below as a quick summary we import the generative model we Define the model name we initiated the chat and send the question to the chat model now I'm going to run this code the question is what is the stock price of Apple so in your terminal python function calling. and click enter and here is the response as AA language model I don't have real time access to the stock market and cannot provide the stock price of Apple this is when the function calling come in you can integrate your own application you can integrate your own function in Gemini how can we do that in this image we can see already we have initiated our model and we are sending prompt to gerini model but now we are going to add extra one more step which is function calling we're going to install why Finance this is required to get the realtime stock price so in this we are going to create application to get the realtime stock price and integrate that in gerini using function calling so as a first step you are importing y Finance as YF next we are going to add function declaration part and Tool these are required to perform function calling and next we're going to write the get stock price function this function get the ticker symbol and then use that to get the latest stock price next we need to write function definition so I'm going to add a tools variable the function is tool which we have already declared at the top and this is the way you define the function function declarations and sending the list of functions so here we are passing only one function which is get stock price and we are defining the properties ticker is the only parameter for example by providing AAPL it will automatically give us the stock price of Apple now we're going to add this Tools in the generative model function now we have completed model initialization and sending a prompt the third step is to check the function calling in this we are getting the function calling from the response we received now we're going to print function call and also let's see the response I'm going to delete this in your terminal python function calling. pi and click enter and here is the response so we receive the function call with the Apple symbol AAPL and the full response now we need to pause this run the function and send the value back to Gemini in this step we are checking if there is a function calling happening or not if it is happening then we are processing that and sending back the response to Gemini let's try to write that in coding so here we are going to add the list of functions which is get stock price and check it against the received response if the received function name contains get stock price then get the arguments that is a symbol and then run the function using this this will automatically run the get stock price function retrieve the stock price of Apple and now we need to send that back to Gemini so to do that response is equals chat. send message we are using the partt from function response and sending the function name that is get stock price and the and also the value which we received by running the get stock function finally chat response we getting the value and then printing the response else we are printing the chat response without function calling so here we check if a function calling is initiated if it is initiated then it will automatically run the function send the result back to gerini IF function calling is not initiated then it will automatically print the default response which we received from step two as we have seen here initially we are sending the prompt if a function calling is initiated then it will come to function calling and then print the results IF function calling is not initiated it will directly print the results as a quick summary we created get stock price function this will automatically get the stock price using Yahoo finance then we are defining the tools and passing this tools to the generative model tools next we are checking if function calling is initiated by Gemini if it is initiated then run the function and send the response back to gerini if no function calling is initiated just print the original response that's it now we going to run this code in your terminal python function calling. pi and click enter so we can see function calling got initiated and we got the answer the current stock price of Apple is $11 19757 as simple as that now you are able to integrate your own function own program into Gemini I'm going to create more videos in regards to Gemini so stay tuned I hope you like this video do like share and subscribe and thanks for watching