Skip to content

Releases: ImahnShekhzadeh/lstm_vision

1.4.0

19 May 16:18
Compare
Choose a tag to compare
  • Fix DDP (on 2 RTXA4000 Ada, runtime compared to a single RTX 4090 is decreased by at least $25 %$.
  • Use hydra both as orchestration tool and for setting up the logs.
  • Remove setup of manual logging, now that hydra is used.
  • Simplify calculation of losses, use as loss summed categorical cross entropy and then take mean for backprop.

1.3.3

16 May 21:47
Compare
Choose a tag to compare

Add logging of energy consumption of entire training.

Fix Worker Seeds in DataLoader

13 May 21:11
Compare
Choose a tag to compare

v1.3.1

27 Jan 18:04
Compare
Choose a tag to compare

Use of wandb for logging now possible

1.3.0

25 Jan 03:11
Compare
Choose a tag to compare

New feature: Training via DDP possible now. If flag --use_ddp is specified and there are at least two GPUs, then DDP will automatically be used.

torch.compile()

07 Jan 16:08
Compare
Choose a tag to compare
  • Highlight: Optimized (compiled) LSTM can be used by using the flag --compile_mode and specifying a compilation mode.
  • Other changes include refactoring of the entire code base (e.g. by moving the code for training and validation into a separate function).
  • It is also way easier now to use another CV dataset than MNIST, check the README for this!

Half-precision

21 Dec 00:29
Compare
Choose a tag to compare

Using half-precision (i.e. torch.float16 instead of torch.float32) greatly reduces the memory consumption, while the necessary code modifications are minor!

LSTM on MNIST

21 Dec 00:14
Compare
Choose a tag to compare

This code contains the code to train a bidirectional LSTM on the MNIST data. With some simple modifications, any dataset can be taken.