-
Notifications
You must be signed in to change notification settings - Fork 1
/
run_long_n2c2_2018_lr5e-5.sh
215 lines (201 loc) · 6.47 KB
/
run_long_n2c2_2018_lr5e-5.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
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
export DATASET_NAME=n2c2_2018_track1
export TOKENIZERS_PARALLELISM=false
# # Baselines (bert-base-uncased, bert-base-cased, PubmedBERT, ClinicalBERT)
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bert-base-uncased \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 512 \
# --per_device_train_batch_size 8 \
# --gradient_accumulation_steps 1 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bert-base-uncased-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bert-base-cased \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 512 \
# --per_device_train_batch_size 8 \
# --gradient_accumulation_steps 1 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bert-base-cased-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 512 \
# --per_device_train_batch_size 8 \
# --gradient_accumulation_steps 1 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/pubmed-bert-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path emilyalsentzer/Bio_ClinicalBERT \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 512 \
# --per_device_train_batch_size 8 \
# --gradient_accumulation_steps 1 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/clinical-bert-$DATASET_NAME/ \
# --overwrite_output_dir
# # Bio-LM
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bio-lm \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 514 \
# --per_device_train_batch_size 8 \
# --gradient_accumulation_steps 1 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bio-lm-514-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bio-lm \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 1028 \
# --per_device_train_batch_size 4 \
# --gradient_accumulation_steps 2 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bio-lm-1028-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bio-lm \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 2056 \
# --per_device_train_batch_size 4 \
# --gradient_accumulation_steps 2 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bio-lm-2056-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bio-lm \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 4112 \
# --per_device_train_batch_size 2 \
# --gradient_accumulation_steps 4 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bio-lm-4112-$DATASET_NAME/ \
# --overwrite_output_dir
# CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
# --model_name_or_path bio-lm \
# --dataset_name $DATASET_NAME \
# --do_train \
# --do_eval \
# --logging_steps 10 \
# --max_seq_length 8224 \
# --per_device_train_batch_size 1 \
# --gradient_accumulation_steps 8 \
# --learning_rate 5e-5 \
# --num_train_epochs 80 \
# --evaluation_strategy epoch \
# --output_dir ./save/bio-lm-9224-$DATASET_NAME/ \
# --overwrite_output_dir
# Bioelectra
CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
--model_name_or_path kamalkraj/bioelectra-base-discriminator-pubmed \
--dataset_name $DATASET_NAME \
--do_train \
--do_eval \
--logging_steps 10 \
--max_seq_length 512 \
--per_device_train_batch_size 8 \
--gradient_accumulation_steps 1 \
--learning_rate 5e-5 \
--num_train_epochs 80 \
--evaluation_strategy epoch \
--output_dir ./save/bioelectra-512-lr5e-5-$DATASET_NAME/ \
--overwrite_output_dir
CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
--model_name_or_path kamalkraj/bioelectra-base-discriminator-pubmed \
--dataset_name $DATASET_NAME \
--do_train \
--do_eval \
--logging_steps 10 \
--max_seq_length 1024 \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 2 \
--learning_rate 5e-5 \
--num_train_epochs 80 \
--evaluation_strategy epoch \
--output_dir ./save/bioelectra-1024-lr5e-5-$DATASET_NAME/ \
--overwrite_output_dir
CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
--model_name_or_path kamalkraj/bioelectra-base-discriminator-pubmed \
--dataset_name $DATASET_NAME \
--do_train \
--do_eval \
--logging_steps 10 \
--max_seq_length 2048 \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 2 \
--learning_rate 5e-5 \
--num_train_epochs 80 \
--evaluation_strategy epoch \
--output_dir ./save/bioelectra-2048-lr5e-5-$DATASET_NAME/ \
--overwrite_output_dir
CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
--model_name_or_path kamalkraj/bioelectra-base-discriminator-pubmed \
--dataset_name $DATASET_NAME \
--do_train \
--do_eval \
--logging_steps 10 \
--max_seq_length 4096 \
--per_device_train_batch_size 2 \
--gradient_accumulation_steps 4 \
--learning_rate 5e-5 \
--num_train_epochs 80 \
--evaluation_strategy epoch \
--output_dir ./save/bioelectra-4096-lr5e-5-$DATASET_NAME/ \
--overwrite_output_dir
CUDA_VISIBLE_DEVICES=0 python run_nlu.py \
--model_name_or_path kamalkraj/bioelectra-base-discriminator-pubmed \
--dataset_name $DATASET_NAME \
--do_train \
--do_eval \
--logging_steps 10 \
--max_seq_length 8192 \
--per_device_train_batch_size 1 \
--gradient_accumulation_steps 8 \
--learning_rate 5e-5 \
--num_train_epochs 80 \
--evaluation_strategy epoch \
--output_dir ./save/bioelectra-8192-lr5e-5-$DATASET_NAME/ \
--overwrite_output_dir