zc/mlp-project
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

FEATURE: export to DataFrame in hdf5

Browse source
This commit is contained in:
Zhengyi Chen 2024-03-06 02:37:18 +00:00
parent 208091ce8a
commit ae9bc34fde
2 changed files with 35 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

4
arguments.py
View file

@ -79,6 +79,10 @@ parser.add_argument(
parser.add_argument( parser.add_argument(
"--debug", type=bool, default=False "--debug", type=bool, default=False
) )
parser.add_argument(
"--export_to_h5", type=bool, default=True,
help="Export training & validation statistics (.h5 of pd.DataFrame)"
)
# nni configuration ========================================================== # nni configuration ==========================================================
parser.add_argument( parser.add_argument(

37
train.py
View file

@ -14,6 +14,7 @@ import torchvision
import nni import nni
import logging import logging
import numpy as np import numpy as np
import pandas as pd
from model.transcrowd_gap import VisionTransformerGAP from model.transcrowd_gap import VisionTransformerGAP
from arguments import args, ret_args from arguments import args, ret_args
@ -23,7 +24,15 @@ from model.transcrowd_gap import *
from checkpoint import save_checkpoint from checkpoint import save_checkpoint
logger = logging.getLogger("train") logger = logging.getLogger("train")
writer = SummaryWriter(args.save_path + "/tensorboard-run")
if not args.export_to_h5:
writer = SummaryWriter(args.save_path + "/tensorboard-run")
else:
train_df = pd.DataFrame(columns=["l1loss", "composite-loss"])
train_stat_file = args.save_path + "/train_stats.h5"
test_df = pd.DataFrame(columns=["mse", "mae"])
test_stat_file = args.save_path + "/test_stats.h5"
def setup_process_group( def setup_process_group(
rank: int, rank: int,
@ -264,7 +273,12 @@ def train_one_epoch(
out, gt_count = model(img, kpoint) out, gt_count = model(img, kpoint)
# loss # loss
loss = criterion(out, gt_count) # wrt. transformer loss = criterion(out, gt_count) # wrt. transformer
writer.add_scalar("L1-loss wrt. xformer (train)", loss, epoch * i) if args.export_to_h5:
train_df.loc[epoch * i, "l1loss"] = loss.item()
else:
writer.add_scalar(
"L1-loss wrt. xformer (train)", loss, epoch * i
)
loss += ( loss += (
F.mse_loss( # stn: info retainment F.mse_loss( # stn: info retainment
@ -276,7 +290,10 @@ def train_one_epoch(
value=loss.item() value=loss.item()
) )
) )
writer.add_scalar("Composite loss (train)", loss, epoch * i) if args.export_to_h5:
train_df.loc[epoch * i, "composite-loss"] = loss.item()
else:
writer.add_scalar("Composite loss (train)", loss, epoch * i)
# free grad from mem # free grad from mem
optimizer.zero_grad(set_to_none=True) optimizer.zero_grad(set_to_none=True)
@ -291,7 +308,10 @@ def train_one_epoch(
break break
# Flush writer # Flush writer
writer.flush() if args.export_to_h5:
train_df.to_hdf(train_stat_file, key="df", mode="a", append=True)
else:
writer.flush()
scheduler.step() scheduler.step()
@ -346,8 +366,13 @@ def valid_one_epoch(test_loader, model, device, epoch, args):
mae = mae * 1.0 / (len(test_loader) * batch_size) mae = mae * 1.0 / (len(test_loader) * batch_size)
mse = np.sqrt(mse / (len(test_loader)) * batch_size) mse = np.sqrt(mse / (len(test_loader)) * batch_size)
writer.add_scalar("MAE (valid)", mae, epoch) if args.export_to_h5:
writer.add_scalar("MSE (valid)", mse, epoch) test_df.loc[epoch, "mae"] = mae
test_df.loc[epoch, "mse"] = mse
test_df.to_hdf(test_stat_file, key="df", mode="a", append=True)
else:
writer.add_scalar("MAE (valid)", mae, epoch)
writer.add_scalar("MSE (valid)", mse, epoch)
if len(xformed) != 0: if len(xformed) != 0:
img_grid = torchvision.utils.make_grid(xformed) img_grid = torchvision.utils.make_grid(xformed)
writer.add_image("STN: transformed image", img_grid, epoch) writer.add_image("STN: transformed image", img_grid, epoch)