What do you think about training custom models on Vertex AI? Let us know in the comments and subscribe for more Prototype to Production episodes → goo.gle/GoogleCloudTech
Como es posible que esta info no esté en español, acabo de ver que es funcional hace mucho y apenas me enteró todo lo que hubiera podido hacer si lo hubiera conocido antes... Algunas veces si es demasiado tarde... Igual me alegra haber encontrado esto
Is vertex custom training done on our local computer? In other words, does vertex custom training allow us to train our model using our local computing, and ram on our personal computer instead of Google cloud?
Very interesting, I would like to know if you could retrain an ai for text creation (write a novel with my own writing style)? Right now what is attracting more attention is AI development in image creation, training to adapt or create content based on a new or mixed style, etc. It seems that text development has lagged behind, compared to the boom that is happening with, for example, Stable Diffusion.
I can not use the /gcs/MY_BUCKET/ path. My service account has the "Storage Object Admin" Role and I could mount it manually to a given folder. But I cant use it by default as shown here. The bucket is in the same location as the notebook and in the same project. Did you skip a necessary step here? Edit: never mind. This only works in the training pipeline but when you want to test initially in the notebook you need to mount a folder manually
@@roopeshk.r3219 its been a while, but I think I used this: cloud.google.com/blog/topics/developers-practitioners/cloud-storage-file-system-vertex-ai-workbench-notebooks?hl=en