Try to reduce memory consumption (experimental)

This commit is contained in:
litagin02
2024-02-07 00:12:38 +09:00
parent cae019d401
commit a0d4444db7

View File

@@ -192,13 +192,13 @@ def run():
train_dataset, train_dataset,
# num_workers=min(config.train_ms_config.num_workers, os.cpu_count() - 1), # num_workers=min(config.train_ms_config.num_workers, os.cpu_count() - 1),
# Slow and often freezes, so use only half of the cores. # Slow and often freezes, so use only half of the cores.
num_workers=min(config.train_ms_config.num_workers, os.cpu_count() // 2), # num_workers=min(config.train_ms_config.num_workers, os.cpu_count() // 2),
shuffle=False, shuffle=False,
pin_memory=True, pin_memory=True,
collate_fn=collate_fn, collate_fn=collate_fn,
batch_sampler=train_sampler, batch_sampler=train_sampler,
persistent_workers=True, persistent_workers=True,
prefetch_factor=6, # prefetch_factor=6,
) # DataLoader config could be adjusted. ) # DataLoader config could be adjusted.
eval_dataset = None eval_dataset = None
eval_loader = None eval_loader = None
@@ -307,16 +307,28 @@ def run():
) )
else: else:
optim_wd = None optim_wd = None
net_g = DDP(net_g, device_ids=[local_rank], bucket_cap_mb=512) net_g = DDP(
net_d = DDP(net_d, device_ids=[local_rank], bucket_cap_mb=512) net_g,
device_ids=[local_rank],
# bucket_cap_mb=512
)
net_d = DDP(
net_d,
device_ids=[local_rank],
# bucket_cap_mb=512
)
if net_dur_disc is not None: if net_dur_disc is not None:
net_dur_disc = DDP( net_dur_disc = DDP(
net_dur_disc, net_dur_disc,
device_ids=[local_rank], device_ids=[local_rank],
bucket_cap_mb=512, # bucket_cap_mb=512,
) )
if net_wd is not None: if net_wd is not None:
net_wd = DDP(net_wd, device_ids=[local_rank], bucket_cap_mb=512) net_wd = DDP(
net_wd,
device_ids=[local_rank],
# bucket_cap_mb=512
)
if utils.is_resuming(model_dir): if utils.is_resuming(model_dir):
if net_dur_disc is not None: if net_dur_disc is not None: