forked from IndoNLP/indonlu
-
Notifications
You must be signed in to change notification settings - Fork 0
/
run_single_task.sh
executable file
·72 lines (51 loc) · 9.37 KB
/
run_single_task.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
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
# albert-base-uncased-96000
# CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-uncased-96000 --step_size 1 --gamma 0.9 --experiment_name albert-base-uncased-96000_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# # albert-base-uncased-96000-spm
# CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-uncased-96000-spm --step_size 1 --gamma 0.9 --experiment_name albert-base-uncased-96000-spm_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# # albert-base-uncased-112500-spm
# CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-uncased-112500-spm --step_size 1 --gamma 0.9 --experiment_name albert-base-uncased-112500-spm_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# scratch
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint scratch --step_size 1 --gamma 0.9 --experiment_name scratch_b$4_step1_gamma0.9_lr1e-4_early$3_layer2_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 2 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint scratch --step_size 1 --gamma 0.9 --experiment_name scratch_b$4_step1_gamma0.9_lr1e-4_early$3_layer4_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 4 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint scratch --step_size 1 --gamma 0.9 --experiment_name scratch_b$4_step1_gamma0.9_lr1e-4_early$3_layer6_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 6 --force
# fasttext-cc-id-300-no-oov-uncased
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-cc-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-cc-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer2_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 2 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-cc-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-cc-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer4_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 4 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-cc-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-cc-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer6_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 6 --force
# fasttext-4B-id-300-no-oov-uncased
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-4B-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-4B-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer2_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 2 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-4B-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-4B-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer4_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 4 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint fasttext-4B-id-300-no-oov-uncased --step_size 1 --gamma 0.9 --experiment_name fasttext-4B-id-300-no-oov-uncased_b$4_step1_gamma0.9_lr1e-4_early$3_layer6_lowerTrue --lr 1e-4 --early_stop $3 --dataset $2 --lower --num_layers 6 --force
# babert-base-512
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-base-512 --step_size 1 --gamma 0.9 --experiment_name babert-base-512_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# babert-bpe-mlm-large-512
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-bpe-mlm-large-512 --step_size 1 --gamma 0.9 --experiment_name babert-bpe-mlm-large-512_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# mbert
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint bert-base-multilingual-uncased --step_size 1 --gamma 0.9 --experiment_name bert-base-multilingual-uncased_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerFalse --lr 1e-5 --early_stop $3 --dataset $2 --num_layers 12 --force
# xlm-roberta
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint xlm-roberta-base --step_size 1 --gamma 0.9 --experiment_name xlm-roberta-base_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerFalse --lr 1e-5 --early_stop $3 --dataset $2 --num_layers 12 --force
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint xlm-roberta-large --step_size 1 --gamma 0.9 --experiment_name xlm-roberta-large_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerFalse --lr 1e-5 --early_stop $3 --dataset $2 --num_layers 24 --force
# babert-opensubtitle
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-opensubtitle --step_size 1 --gamma 0.9 --experiment_name babert-opensubtitle_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerFalse --lr 1e-5 --early_stop $3 --dataset $2 --num_layers 12 --force
# xlm
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint xlm-mlm-100-1280 --step_size 1 --gamma 0.9 --experiment_name xlm-mlm-100-1280_b$4_step1_gamma0.9_lr1e-5_early$3_layer16_lowerFalse --lr 1e-5 --early_stop $3 --dataset $2 --num_layers 16 --force
# albert-large-wwmlm-128
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-large-wwmlm-128 --step_size 1 --gamma 0.9 --experiment_name albert-large-wwmlm-128_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# albert-base-wwmlm-512
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-wwmlm-512 --step_size 1 --gamma 0.9 --experiment_name albert-base-wwmlm-512_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# albert-large-wwmlm-512
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-large-wwmlm-512 --step_size 1 --gamma 0.9 --experiment_name albert-large-wwmlm-512_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# albert-base-uncased-112500
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-uncased-112500 --step_size 1 --gamma 0.9 --experiment_name albert-base-uncased-112500_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# albert-base-uncased-191k
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint albert-base-uncased-191k --step_size 1 --gamma 0.9 --experiment_name albert-base-uncased-191k_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# cartobert
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint cartobert --step_size 1 --gamma 0.9 --experiment_name cartobert_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force
# babert-bpe-mlm-large-uncased
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-bpe-mlm-large-uncased --step_size 1 --gamma 0.9 --experiment_name babert-bpe-mlm-large-uncased_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# babert-bpe-mlm-large-uncased-1m
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-bpe-mlm-large-uncased-1m --step_size 1 --gamma 0.9 --experiment_name babert-bpe-mlm-large-uncased-1m_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# babert-bpe-mlm-large-uncased-1100k
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-bpe-mlm-large-uncased-1100k --step_size 1 --gamma 0.9 --experiment_name babert-bpe-mlm-large-uncased-1100k_b$4_step1_gamma0.9_lr1e-5_early$3_layer24_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 24 --force
# babert-bpe-mlm-uncased-128-dup10-5
CUDA_VISIBLE_DEVICES=$1 python3 main.py --n_epochs 100 --train_batch_size $4 --model_checkpoint babert-bpe-mlm-uncased-128-dup10-5 --step_size 1 --gamma 0.9 --experiment_name babert-bpe-mlm-uncased-128-dup10-5_b$4_step1_gamma0.9_lr1e-5_early$3_layer12_lowerTrue --lr 1e-5 --early_stop $3 --dataset $2 --lower --num_layers 12 --force