Skip to content

fix: resuming a snapshot *.pkl file takes into consideration kimg value. #82

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

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

fix: resuming a snapshot *.pkl file takes into consideration kimg value. #82

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
0e9b616
fix: 'resume' from a snapshot continue from latest trained kimg.
mb0rt Apr 1, 2021
464cef9
fix: initial progress when resuming a training.
mb0rt Apr 27, 2021
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
12 changes: 10 additions & 2 deletions training/training_loop.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
# license agreement from NVIDIA CORPORATION is strictly prohibited.

import os
import re
import time
import copy
import json
Expand Down Expand Up @@ -152,13 +153,20 @@ def training_loop(
G_ema = copy.deepcopy(G).eval()

# Resume from existing pickle.
cur_nimg = 0
if (resume_pkl is not None) and (rank == 0):
print(f'Resuming from "{resume_pkl}"')
with dnnlib.util.open_url(resume_pkl) as f:
resume_data = legacy.load_network_pkl(f)
for name, module in [('G', G), ('D', D), ('G_ema', G_ema)]:
misc.copy_params_and_buffers(resume_data[name], module, require_all=False)

# resume from a snapshot file(`network-snapshot-<INT>.pkl`) continues from
# where it stopped..
match = re.match(r"^.*(network-snapshot-)(\d+)(.pkl)$", resume_pkl, re.IGNORECASE)
if match:
cur_nimg = int(match.group(2)) * 1000

# Print network summary tables.
if rank == 0:
z = torch.empty([batch_gpu, G.z_dim], device=device)
Expand Down Expand Up @@ -245,14 +253,14 @@ def training_loop(
if rank == 0:
print(f'Training for {total_kimg} kimg...')
print()
cur_nimg = 0

cur_tick = 0
tick_start_nimg = cur_nimg
tick_start_time = time.time()
maintenance_time = tick_start_time - start_time
batch_idx = 0
if progress_fn is not None:
progress_fn(0, total_kimg)
progress_fn(cur_nimg // 1000, total_kimg)
while True:

# Fetch training data.
Expand Down