UniqAI/gaussian-splatting
1
0
Fork 0
mirror of https://github.com/graphdeco-inria/gaussian-splatting synced 2025-06-26 18:18:11 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
Ammar Qammaz 2023-11-03 09:31:20 +02:00
parent af89ad77ce
commit 4e36f28d47
1 changed files with 9 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
train.py
View File

@ -29,8 +29,9 @@ except ImportError:
TENSORBOARD_FOUND = False
def training(dataset, opt, pipe, testing_iterations, saving_iterations, checkpoint_iterations, checkpoint, debug_from):
bestLossEncountered = 1000000
pickBetweenFinalNLosses = 10
bestLossEncountered = 1000000
bestIterationEncountered = 0
pickBetweenFinalNLosses = 10
first_iter = 0
tb_writer = prepare_output_and_logger(dataset)
@ -107,7 +108,12 @@ def training(dataset, opt, pipe, testing_iterations, saving_iterations, checkpoi
progress_bar.close()
if (bestLossEncountered>ema_loss_for_log) and (pickBetweenFinalNLosses+iteration > opt.iterations):
print("\n[ITER {}] Also remembering this iteration..".format(iteration))
if (bestIterationEncountered!=0):
print("\n[ITER {}] Erasing this iteration..".format(bestIterationEncountered))
os.system("rm point_cloud/iteration_%u/*.ply && rmdir point_cloud/iteration_%u/"%iteration)
bestLossEncountered = ema_loss_for_log
bestIterationEncountered = iteration
print("\n[ITER {}] Now remembering this iteration..".format(iteration))
saving_iterations.append(iteration)
# Log and save