Thanks for this video! I have no knowledge of AI/ML. I just want to use vertex ai for my mobile application purposes. Now I have a dataset in the Kaggle. I have downloaded the dataset and all are the images. Now I want to use that to create an model and use API into my mobile app. How to do it in the GUI of vertex AI? I mean there are only images. As there isn't any CSV or JSON file, it's very time-consuming to upload 100 of images and label every one because I will be using Image Object detection. Is there any direct way to get the CSV or JSON from kaggle. or Can I get a trained model direclty. what's the exact flow to do get what I want?
Hi, you need to prepare either a CSV or JSONL file with image locations and labels, once you have it, use it to create an image dataset in Vertex AI, this, of course, requires some programming knowledge as I don't believe there will be a ready file for that purpose on Kaggle. I'm showing the process of creating such file from images I uploaded to Cloud Storage, hopefully, it can be a good start for you. If you have any doubts regarding the file, you can always refer to the documentation => cloud.google.com/vertex-ai/docs/image-data/classification/prepare-data
At what stage are you getting this error? What I can recommend from the top of my head is to try to authenticate with GCP in the similar way as I do it on this video: th-cam.com/video/U0Rrqgid6KU/w-d-xo.html&ab_channel=Cloud4DataScience
Yes, generally you should create your input_file.csv file in a way that contains the addresses of all images that you want to include inside the dataset. Your CSV file will have the address in the 1st column, and the labels of images in the 2nd column. There is no need to create multiple CSV files, just this 1 with all the addresses and labels, hope it helps :)
@@nityarakhe2351 Hi, you need to get a name of the current bucket that you are currently listing files from. Most probably you can do it either with Python or with Google Cloud SDK (gcloud storage …).
@@nityarakhe2351 I apologize but I cannot provide direct help with the code you are working on. I'm afraid I don't have the time needed to do this. Also, I am sure that you will be able to figure this out by finding other resources on the internet that will help you to achieve it.
Hello, I am discussing the process of creation of input_file.csv in the "Create CSV and JSONL input file" chapter. It comes down to creating a simple file with 2 columns where the first one has addresses of images inside the Cloud Storage and the second has the label. This is an example of such a file: github.com/rafaello9472/c4ds/blob/main/Create%20image%20dataset%20in%20Vertex%20AI/input_file.csv Then you need to store this file inside the Cloud Storage and use it as input for creating an image dataset. If something is still unclear please let me know, I will try to help :)
Thanks for this video! I have no knowledge of AI/ML. I just want to use vertex ai for my mobile application purposes. Now I have a dataset in the Kaggle. I have downloaded the dataset and all are the images. Now I want to use that to create an model and use API into my mobile app. How to do it in the GUI of vertex AI? I mean there are only images. As there isn't any CSV or JSON file, it's very time-consuming to upload 100 of images and label every one because I will be using Image Object detection. Is there any direct way to get the CSV or JSON from kaggle. or Can I get a trained model direclty. what's the exact flow to do get what I want?
Hi, you need to prepare either a CSV or JSONL file with image locations and labels, once you have it, use it to create an image dataset in Vertex AI, this, of course, requires some programming knowledge as I don't believe there will be a ready file for that purpose on Kaggle. I'm showing the process of creating such file from images I uploaded to Cloud Storage, hopefully, it can be a good start for you. If you have any doubts regarding the file, you can always refer to the documentation => cloud.google.com/vertex-ai/docs/image-data/classification/prepare-data
Awesome video! I have a problem when uploading my csv or JSON file. It says Warning annotations deduped. Ever encountered this?
Yes, the possible cause is that there are duplicates in the input file or the file format is wrong.
@@cloud4datascience772 thanks a lot!!
Thanks for the video. I keep getting a '401 Request had invalid authentication credentials ' how can i solve this?
At what stage are you getting this error? What I can recommend from the top of my head is to try to authenticate with GCP in the similar way as I do it on this video: th-cam.com/video/U0Rrqgid6KU/w-d-xo.html&ab_channel=Cloud4DataScience
instead of merging all individual csv file is there any other way to get all images paths in one file
Yes, generally you should create your input_file.csv file in a way that contains the addresses of all images that you want to include inside the dataset. Your CSV file will have the address in the 1st column, and the labels of images in the 2nd column. There is no need to create multiple CSV files, just this 1 with all the addresses and labels, hope it helps :)
@@nityarakhe2351 Hi, you need to get a name of the current bucket that you are currently listing files from. Most probably you can do it either with Python or with Google Cloud SDK (gcloud storage …).
@@nityarakhe2351 I apologize but I cannot provide direct help with the code you are working on. I'm afraid I don't have the time needed to do this. Also, I am sure that you will be able to figure this out by finding other resources on the internet that will help you to achieve it.
whats is gcs_source?
how your given input_file.csv, how your created
Hello, I am discussing the process of creation of input_file.csv in the "Create CSV and JSONL input file" chapter. It comes down to creating a simple file with 2 columns where the first one has addresses of images inside the Cloud Storage and the second has the label. This is an example of such a file: github.com/rafaello9472/c4ds/blob/main/Create%20image%20dataset%20in%20Vertex%20AI/input_file.csv
Then you need to store this file inside the Cloud Storage and use it as input for creating an image dataset. If something is still unclear please let me know, I will try to help :)
@@cloud4datascience772 Thanks