Feature: Add Support for Choosing GPU Index about hub HOT 11 CLOSED

@sergiuwaxmann This is something we need to ensure is possible with the new advanced options. I have seen this problem regularly in the wild. Currently we assign the GPU index incrementally from 0. HUB users need the option to overwrite this with an array of custom indexes.

from hub.

👋 Hello @Pedro-Leitek, thank you for raising an issue about Ultralytics HUB 🚀! Please visit our HUB Docs to learn more:

• Quickstart. Start training and deploying YOLO models with HUB in seconds.
• Datasets: Preparing and Uploading. Learn how to prepare and upload your datasets to HUB in YOLO format.
• Projects: Creating and Managing. Group your models into projects for improved organization.
• Models: Training and Exporting. Train YOLOv5 and YOLOv8 models on your custom datasets and export them to various formats for deployment.
• Integrations. Explore different integration options for your trained models, such as TensorFlow, ONNX, OpenVINO, CoreML, and PaddlePaddle.
• Ultralytics HUB App. Learn about the Ultralytics App for iOS and Android, which allows you to run models directly on your mobile device.
  • iOS. Learn about YOLO CoreML models accelerated on Apple's Neural Engine on iPhones and iPads.
  • Android. Explore TFLite acceleration on mobile devices.
• Inference API. Understand how to use the Inference API for running your trained models in the cloud to generate predictions.

If this is a 🐛 Bug Report, please provide screenshots and steps to reproduce your problem to help us get started working on a fix.

If this is a ❓ Question, please provide as much information as possible, including dataset, model, environment details etc. so that we might provide the most helpful response.

We try to respond to all issues as promptly as possible. Thank you for your patience!

from hub.

@Pedro-Leitek We would need more information to be helpful here. The model id that is giving you problems if trained from HUB or the full command if you are using the ultralytics package directly.

from hub.

@Pedro-Leitek We would need more information to be helpful here. The model id that is giving you problems if trained from HUB or the full command if you are using the ultralytics package directly.

@kalenmike ok, the model is yolov5m6u, the command I'm using is the following:

from ultralytics import YOLO, checks, hub
checks()

hub.login('')

model = YOLO('https://hub.ultralytics.com/models/lvOqRZ7rWGdek3K5Mdwk')
results = model.train()

from hub.

@Pedro-Leitek The model you have set here is a single GPU model?

Also note that sharing your API key will allow other users access to your account.

from hub.

@kalenmike yes it was, here is the code for multi-gpu

from ultralytics import YOLO, checks, hub
checks()

hub.login('')

model = YOLO('https://hub.ultralytics.com/models/lvOqRZ7rWGdek3K5Mdwk')
results = model.train()

from hub.

@Pedro-Leitek Do you have logs for the server you were training on? One issue could be that you are assigned GPU's that are at a different index. You are currently requesting GPU 0 and 1, perhaps they are already busy or they are not available?

Also ensure that you have access to the GPU from where you are running the code.

from hub.

@kalenmike I read somewhere that it may work if I request GPU 0, 2 or 1,3. For some reason it worked with another user. I didn't notice if the gpus were consecutive. But, lets just say I rent a machine that the gpus are not consecutive, lets say 0,2, how can I force the hub to set devices 0, 2 (or any other sequence)?

from hub.

@Pedro-Leitek Currently the device id's cannot be set on HUB for multi GPU training. We are planning to release more advanced configuration soon. The solution at the moment is to train the model from your server and connect it to HUB by logging in first, this will mean that your model will not have a connection to the dataset.

If you want to go this route here is an example, you will need to have your dataset available on the server and pass all the arguments to model.train.

from ultralytics import YOLO, checks, hub
checks()

hub.login(API_KEY)

model = YOLO('yolov8n.pt')
results = model.train(TRAIN_ARGS)

from hub.

👋 Hello there! We wanted to give you a friendly reminder that this issue has not had any recent activity and may be closed soon, but don't worry - you can always reopen it if needed. If you still have any questions or concerns, please feel free to let us know how we can help.

For additional resources and information, please see the links below:

• Docs: https://docs.ultralytics.com
• HUB: https://hub.ultralytics.com
• Community: https://community.ultralytics.com

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLO 🚀 and Vision AI ⭐

from hub.

@Pedro-Leitek Hello!
First of all, my apologies for the delayed response.
I'm pleased to inform you that we introduced a new "Custom" option in the Advanced Model Configuration approximately two weeks ago. This enhancement allows users to utilize all supported training settings. It also includes the capability to specify custom indexes (see image attached below).

from hub.