could you share what's hardware(GPU) used to train this model?

JackieHanLab / TOSICA

Transformer for One-Stop Interpretable Cell-type Annotation

MIT License

121 stars 23 forks source link

could you share what's hardware(GPU) used to train this model? #7

Closed rushrush2022 closed 1 year ago

rushrush2022 commented 1 year ago

@JackieHanLab

thanks very much for you to share the codes, could you share what's hardware(mostly GPU) configuration used to train these models mentioned in paper? what's the time used ?

thanks again!!

rushrush2022 commented 1 year ago

From tutorial.ipynb, I see "NVIDIA GeForce RTX 3080 Ti Laptop GPU", yes? thanks

JiaweiChenGo commented 1 year ago

In single Tesla V100-SXM2-32GB: train 85~90 cells/s and predict ~2500 cells/s; In single RTX 3080 Ti Laptop GPU 16G: train ~70 cells/s and predict ~2000 cells/s