-
Notifications
You must be signed in to change notification settings - Fork 0
/
pretraining.sh
45 lines (38 loc) · 1.59 KB
/
pretraining.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
# Set the path to save checkpoints
OUTPUT_DIR='/home/liullhappy/imageNet/MAE-pytorch-main-ablation/output/PT'
# path to imagenet-1k train set
DATA_PATH='/home/liullhappy/imageNet/train'
# Set the path to save TesnorBoard --log_dir
LOG_DIR='/home/liullhappy/imageNet/MAE-pytorch-main/log_output'
# batch_size can be adjusted according to the graphics card
# OMP_NUM_THREADS=1 python -m torch.distributed.launch --nproc_per_node=2 run_mae_pretraining.py \
# --data_path ${DATA_PATH} \cc
# --mask_ratio 0.75 \
# --model pretrain_mae_base_patch16_224 \
# --batch_size 64 \
# --opt adamw \
# --opt_betas 0.9 0.95 \
# --warmup_epochs 40 \
# --epochs 16 \
# --output_dir ${OUTPUT_DIR}
# CUDA_VISIBLE_DEVICES=0 python -m torch.distributed.launch --nproc_per_node=1 --master_port=8675 --use_env run_mae_pretraining.py \
# --data_path ${DATA_PATH} \
# --mask_ratio 0.75 \
# --model pretrain_mae_base_patch16_224 \
# --batch_size 64 \
# --opt adamw \
# --opt_betas 0.9 0.95 \
# --warmup_epochs 40 \
# --epochs 16 \
# --output_dir ${OUTPUT_DIR}
CUDA_VISIBLE_DEVICES=0,1 python -m torch.distributed.launch --nproc_per_node=2 --master_port=8675 --use_env run_mae_pretraining.py \
--data_path ${DATA_PATH} \
--mask_ratio 0.75 \
--model pretrain_mae_base_patch16_224 \
--batch_size 128 \
--opt adamw \
--opt_betas 0.9 0.95 \
--warmup_epochs 40 \
--epochs 100 \
--output_dir ${OUTPUT_DIR} \
--log_dir ${LOG_DIR}