Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add mean loss log #2

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Add mean loss log #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
dab2879
Add mean_loss log
casellimarco Oct 25, 2024
71ef950
Improve printing
casellimarco Oct 28, 2024
fab2809
Cast to the correct device
casellimarco Oct 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 1 addition & 2 deletions pfns/bar_distribution.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -386,8 +386,7 @@ def get_bucket_limits(num_outputs:int, full_range:tuple=None, ys:torch.Tensor=No
full_range = (ys.min(), ys.max())
else:
assert full_range[0] <= ys.min() and full_range[1] >= ys.max(), f'full_range {full_range} not in range of ys {ys.min(), ys.max()}'
# FIXME: this needs to be on same device
full_range = torch.tensor(full_range)
full_range = torch.tensor(full_range, device=ys.device)
ys_sorted, ys_order = ys.sort(0)
bucket_limits = (ys_sorted[ys_per_bucket-1::ys_per_bucket][:-1]+ys_sorted[ys_per_bucket::ys_per_bucket])/2
if verbose:
Expand Down
12 changes: 10 additions & 2 deletions pfns/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,7 @@
import torch
from torch import nn
from torch.cuda.amp import autocast, GradScaler
import numpy as np

from . import utils
from .priors import prior
Expand Down Expand Up @@ -281,6 +282,7 @@ def apply_batch_wise_criterion(i):
try:
total_loss, total_positional_losses, time_to_get_batch, forward_time, step_time, nan_share, ignore_share =\
train_epoch()
step_callback({"mean_loss": total_loss})
except Exception as e:
print("Invalid epoch encountered, skipping...")
print(e)
Expand All @@ -293,10 +295,16 @@ def apply_batch_wise_criterion(i):
val_score = None

if verbose:
pos_losses_str = f"pos losses {','.join([f'{l:5.2f}' for l in total_positional_losses])}"
if len(total_positional_losses) > 20:
actual_losses = np.array(total_positional_losses)[~np.isnan(total_positional_losses)]
pos_losses_str = f"pos losses std {np.std(actual_losses):5.2f} | " + \
f"pos losses quantiles {np.quantile(actual_losses, [0.1, 0.25, 0.5, 0.75, 0.9])} | " + \
f"losses/total {len(actual_losses)} / {len(total_positional_losses)}"
print('-' * 89)
print(
f'| end of epoch {epoch:3d} | time: {(time.time() - epoch_start_time):5.2f}s | mean loss {total_loss:5.2f} | '
f"pos losses {','.join([f'{l:5.2f}' for l in total_positional_losses])}, lr {scheduler.get_last_lr()[0]}"
f'| end of epoch {epoch:3d} | time: {(time.time() - epoch_start_time):5.2f}s | mean loss {total_loss:5.4f} | '
f' {pos_losses_str}, lr {scheduler.get_last_lr()[0]}'
f' data time {time_to_get_batch:5.2f} step time {step_time:5.2f}'
f' forward time {forward_time:5.2f}'
f' nan share {nan_share:5.2f} ignore share (for classification tasks) {ignore_share:5.4f}'
Expand Down