What about the gpu's memory? about segment-anything-finetuner

[nikolaydyankov]

Memory is a big issue for me too, I'm training the smallest model with batch size of 1 with both encoders frozen and it still eats 22GB VRAM. If the model could be trained on 640x640 instead of 1024 images that would be a huge memory saver.

[skyfallsss]

I met some problem.It stops after a few epochs.

[skyfallsss]

The errors are as follow.
/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/callbacks/model_checkpoint.py:636: UserWarning: ModelCheckpoint(monitor='val_per_mask_iou') not found in the returned metrics: ['loss', 'loss_focal', 'loss_dice', 'loss_iou', 'train_per_mask_iou']. HINT: Did you call self.log('val_per_mask_iou', value) in the LightningModule?
warning_cache.warn(m)
Traceback (most recent call last):
File "finetune.py", line 414, in
main()
File "finetune.py", line 398, in main
trainer.fit(model)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 737, in fit
self._call_and_handle_interrupt(
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 682, in _call_and_handle_interrupt
return trainer_fn(*args, **kwargs)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 772, in _fit_impl
self._run(model, ckpt_path=ckpt_path)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 1194, in _run
self._dispatch()
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 1274, in _dispatch
self.training_type_plugin.start_training(self)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/plugins/training_type/training_type_plugin.py", line 202, in start_training
self._results = trainer.run_stage()
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 1284, in run_stage
return self._run_train()
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 1314, in _run_train
self.fit_loop.run()
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/base.py", line 145, in run
self.advance(*args, **kwargs)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/fit_loop.py", line 234, in advance
self.epoch_loop.run(data_fetcher)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/base.py", line 145, in run
self.advance(*args, **kwargs)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py", line 199, in advance
self.update_lr_schedulers("step", update_plateau_schedulers=False)
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py", line 441, in update_lr_schedulers
self._update_learning_rates(
File "/home/kemosheng/.local/lib/python3.8/site-packages/pytorch_lightning/loops/epoch/training_epoch_loop.py", line 505, in _update_learning_rates
lr_scheduler["scheduler"].step()
File "/home/kemosheng/anaconda3/envs/sam/lib/python3.8/site-packages/torch/optim/lr_scheduler.py", line 161, in step
values = self.get_lr()
File "/home/kemosheng/anaconda3/envs/sam/lib/python3.8/site-packages/torch/optim/lr_scheduler.py", line 260, in get_lr
return [base_lr * lmbda(self.last_epoch)
File "/home/kemosheng/anaconda3/envs/sam/lib/python3.8/site-packages/torch/optim/lr_scheduler.py", line 260, in
return [base_lr * lmbda(self.last_epoch)
File "finetune.py", line 299, in warmup_step_lr
if steps >= milestone * self.trainer.estimated_stepping_batches:
AttributeError: 'Trainer' object has no attribute 'estimated_stepping_batches'