-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrun.slurm
235 lines (214 loc) · 6.95 KB
/
run.slurm
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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
#!/bin/sh
#SBATCH --account=boyer
#SBATCH --qos=boyer
#SBATCH --partition=gpu
#SBATCH --gpus=a100:7
#SBATCH -N 1 # nodes requested
#SBATCH -n 1 # tasks requested
#SBATCH -c 2 # cores requested
#SBATCH --mem=60gb # memory in Mb
#SBATCH -o out_cl # send stdout to outfile
#SBATCH -e err_cl # send stderr to errfile
#SBATCH -t 10:00:00 # time requested in hour:minute:second
# Load necessary modules
module load git
module load python/3.8
module load cuda/11.4.3
# Remove any existing old files
# rm -rf continual-learning-nlu
# rm out_cl
# rm err_cl
# rm venv
# Clone the GitHub repository
#git clone https://github.com/msamogh/continual-learning-nlu.git
cd continual-learning-nlu
# git pull
# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate
# Create and activate a virtual environment
pip install -r requirements.txt
# Change directory
cd cl_domain/
# Random Generate dataset + Train + Evaluate
# echo "Strategy = Random"
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode train \
# --cl_super_run_label random-flat-tee \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy random \
# --num_train_epochs 18 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode evaluate \
# --cl_super_run_label random-flat-tee \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy random \
# --num_train_epochs 5 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
# Max path Generate dataset + Train + Evaluate
# echo "Strategy = Max Path"
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode train \
# --cl_super_run_label max_path-vicious-tree \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy max_path \
# --num_train_epochs 18 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode evaluate \
# --cl_super_run_label max_path-vicious-tree \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy max_path \
# --num_train_epochs 5 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
# Min path Generate dataset + Train + Evaluate
echo "Strategy = Min Path"
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode generate_data \
# --cl_super_run_label tahoe-word \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy min_path \
# --num_train_epochs 18 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
--mode train \
--cl_super_run_label max_path-studio-vest \
--cl_checkpoint_dir ../cl_checkpoints \
--cl_run_dir ../cl_runs \
--results_dir ../cl_results \
--ordering_strategy max_path \
--num_train_epochs 18 \
--num_runs 22 \
--num_domains_per_run 5 \
--cl_lr_schedule constant \
--limit_n_samples 180 \
--val_size_per_domain 0.01 \
--cl_experience_replay_size 20 \
--eval_batch_size 4 \
--train_batch_size 4 \
PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
--mode evaluate \
--cl_super_run_label max_path-studio-vest \
--cl_checkpoint_dir ../cl_checkpoints \
--cl_run_dir ../cl_runs \
--results_dir ../cl_results \
--ordering_strategy max_path \
--num_train_epochs 5 \
--num_runs 22 \
--num_domains_per_run 5 \
--cl_lr_schedule constant \
--limit_n_samples 180 \
--val_size_per_domain 0.01 \
--cl_experience_replay_size 20 \
--eval_batch_size 4 \
--train_batch_size 4 \
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode train \
# --cl_super_run_label random-plastic-cover \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy random \
# --num_train_epochs 18 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 4 \
# --train_batch_size 4 \
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode evaluate \
# --cl_super_run_label random-plastic-cover \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy random \
# --num_train_epochs 5 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule constant \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 4 \
# --train_batch_size 4 \
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode all \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy max_path \
# --num_train_epochs 5 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule linear \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \
# # Generate dataset + Train + Evaluate
# echo "Strategy = Min Path"
# PYTHONPATH=$PYTHONPATH:$(pwd)/.. python run.py \
# --mode all \
# --cl_checkpoint_dir ../cl_checkpoints \
# --cl_run_dir ../cl_runs \
# --results_dir ../cl_results \
# --ordering_strategy min_path \
# --num_train_epochs 5 \
# --num_runs 22 \
# --num_domains_per_run 5 \
# --cl_lr_schedule linear \
# --limit_n_samples 180 \
# --val_size_per_domain 0.01 \
# --cl_experience_replay_size 20 \
# --eval_batch_size 16 \
# --train_batch_size 16 \