-
Notifications
You must be signed in to change notification settings - Fork 0
/
eval.sh
45 lines (41 loc) · 1.39 KB
/
eval.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
#!/usr/bin/env bash
set -x # print the commands
export MASTER_PORT=${MASTER_PORT:-12321} # You should set the same master_port in all the nodes
OUTPUT_DIR='Your/path/to/result/dir'
DATA_PATH='Your/path/to/data'
MODEL_PATH='Your/path/to/model'
N_NODES=1 # Number of nodes
GPUS_PER_NODE=1 # Number of GPUs in each node
SRUN_ARGS=${SRUN_ARGS:-""} # Other slurm task args
PY_ARGS=${@:3} # Other training args
# Please refer to `run_class_finetuning.py` for the meaning of the following hyperreferences
CUDA_VISIBLE_DEVICES="3" torchrun --nproc_per_node=1 --master_port=29514 \
run_class_finetuning.py \
--model vit_base_patch16_224 \
--data_set Ekman6 \
--nb_classes 6 \
--data_path ${DATA_PATH} \
--finetune ${MODEL_PATH} \
--log_dir ${OUTPUT_DIR} \
--output_dir ${OUTPUT_DIR} \
--batch_size 1 \
--input_size 224 \
--short_side_size 224 \
--save_ckpt_freq 10 \
--num_frames 16 \
--sampling_rate 4 \
--num_sample 2 \
--num_workers 14 \
--opt adamw \
--lr 3e-4 \
--drop_path 0.3 \
--clip_grad 5.0 \
--layer_decay 0.9 \
--opt_betas 0.9 0.999 \
--weight_decay 0.1 \
--warmup_epochs 0 \
--epochs 50 \
--test_num_segment 12 \
--test_num_crop 3 \
--dist_eval --enable_deepspeed --eval \
${PY_ARGS}