Sayan Biswas

icon indicating a website link https://in.linkedin.com/in/sayan-biswas-867220136 icon indicating an email [email protected]

icon company walmart global tech IDC icon location Bangalore icon location Currently working at Walmart Labs India as a Sr. Data Scientist.

Results 2 comments of Sayan Biswas

Multi GPU Memory keeps increasing while training TFT

@fabsen Thank you for the solution of `log_interval=-1`. I faced the same issue while training in ddp mode on 4x NVidia V100. This was a major hurdle for scalability. Libraries...

Huge Memory Consumption for TFT & Small Dataset

Are you using DDP mode? I think in that case, the `log_interval=10` needs to change to `log_interval=-1`. There's a memory leak if I recall correctly. Issue: [486](https://github.com/jdb78/pytorch-forecasting/issues/486)