I am trying to fine-tune a model for multi-label text classification on my own dataset using the Accelerate module. With a single GPU, everything is OK, but on two GPUs, I am getting the following error: Traceback (most recent call last): File "/home/student/Experiemnts/MultiLabelClassification_L…

[image] akkasi: pred = torch.zeros(outputs['logits'].shape) You need to move this to the GPU

Problem with model inference using accelerate

akkasi November 22, 2023, 2:31pm 4

Thank you @muellerzr the problem is solved.

Topic		Replies	Views
Loading a HF Model in Multiple GPUs and Run Inferences in those GPUs 🤗Accelerate	10	9701	October 16, 2024
How to collect the accuracy when running multi GPU model with accelerate? 🤗Accelerate	3	1005	December 8, 2023
Can't pickle error using accelerate multi-GPU 🤗Accelerate	6	10046	March 7, 2023
Multi node CPU to train transformer GPT-JT-6B-v1 (moved) 🤗Transformers	0	424	February 20, 2023
Accelerator() causes Error Beginners	2	390	April 12, 2024