-
Notifications
You must be signed in to change notification settings - Fork 1
/
prompt-scale.py
177 lines (138 loc) · 5.93 KB
/
prompt-scale.py
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
import argparse
import math
import os
from tqdm import tqdm
from openprompt.data_utils.text_classification_dataset \
import YahooProcessor, SST2Processor, MnliProcessor, AgnewsProcessor
import torch
from openprompt.data_utils.utils import InputExample
from openprompt.plms import load_plm
from openprompt.prompts import ManualTemplate
from openprompt.prompts import ManualVerbalizer, KnowledgeableVerbalizer
from openprompt import PromptForClassification
from openprompt.data_utils.data_sampler import FewShotSampler
from transformers import get_linear_schedule_with_warmup
from openprompt import PromptDataLoader
import torch.multiprocessing as mp
PROCESSER = {
"sst2": SST2Processor,
"mnli": MnliProcessor,
"agnews": AgnewsProcessor,
"yahoo_answers_topics": YahooProcessor,
}
MODEL_PATH = {
"t5-small": "t5-small",
"t5-base": "t5-base",
"t5-large": "t5-large",
"t5-3b": "t5-3b",
"bert-tiny": "bert-tiny",
"bert-mini": "bert-mini",
"bert-small": "bert-small",
"bert-medium": "bert-medium",
"bert-base": "bert-base",
"bert-large": "bert-large",
}
DATASET_PATH = {
"sst2": "./datasets/TextClassification/SST-2",
"mnli": "./datasets/TextClassification/mnli",
"agnews": "./datasets/TextClassification/agnews",
"yahoo_answers_topics": "./datasets/TextClassification/yahoo_answers_topics",
}
NUM_CLASSES = {
"sst2": 2,
"mnli": 3,
"agnews": 4,
"yahoo_answers_topics": 10,
}
import random
import numpy as np
import torch
def set_seed(seed):
print("set seed:", seed)
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)
np.random.seed(seed)
random.seed(seed)
def main(args):
num_classes = args.num_classes
scale = args.scale
model_name = args.model_name
model_path = args.model_path
dataset_name = args.dataset_name
dataset_path = args.dataset_path
seed = args.seed
processer = PROCESSER[dataset_name]()
dataset = {}
dataset['train'] = processer.get_train_examples(dataset_path)
dataset['test'] = processer.get_test_examples(dataset_path)
plm, tokenizer, model_config, WrapperClass = load_plm(model_name.split("-")[0], model_path)
mytemplate = ManualTemplate(tokenizer=tokenizer).from_file(f"scripts/TextClassification/{dataset_name}/manual_template.txt", choice=0)
myverbalizer = ManualVerbalizer(tokenizer, num_classes=num_classes).from_file(f"scripts/TextClassification/{dataset_name}/manual_verbalizer.txt")
prompt_model = PromptForClassification(plm=plm,template=mytemplate, verbalizer=myverbalizer, freeze_plm=False).cuda()
if scale in ["3b", "11b"]:
prompt_model.parallelize()
train_dataloader = PromptDataLoader(dataset=dataset["train"], template=mytemplate, tokenizer=tokenizer,
tokenizer_wrapper_class=WrapperClass, max_seq_length=256, decoder_max_length=3,
batch_size=8,shuffle=True, teacher_forcing=False, predict_eos_token=False,
truncate_method="tail")
test_dataloader = PromptDataLoader(dataset=dataset["test"], template=mytemplate, tokenizer=tokenizer,
tokenizer_wrapper_class=WrapperClass, max_seq_length=256, decoder_max_length=3,
batch_size=8,shuffle=False, teacher_forcing=False, predict_eos_token=False,
truncate_method="head")
loss_func = torch.nn.CrossEntropyLoss()
no_decay = ['bias', 'LayerNorm.weight']
optimizer_grouped_parameters = [
{'params': [p for n, p in prompt_model.named_parameters() if not any(nd in n for nd in no_decay)], 'weight_decay': 0.01},
{'params': [p for n, p in prompt_model.named_parameters() if any(nd in n for nd in no_decay)], 'weight_decay': 0.0}
]
optimizer = torch.optim.AdamW(optimizer_grouped_parameters, lr=1e-5)
prompt_model.train()
for epoch in range(10):
tot_loss = 0
for step, inputs in enumerate(train_dataloader):
inputs = inputs.cuda()
logits = prompt_model(inputs)
labels = inputs['label']
loss = loss_func(logits, labels)
loss.backward()
tot_loss += loss.item()
optimizer.step()
optimizer.zero_grad()
if step %100 == 1:
print("Epoch {}, average loss: {}".format(epoch, tot_loss/(step+1)), flush=True)
# test
prompt_model.eval()
allprobs = []
allpreds = []
alllabels = []
for step, inputs in enumerate(test_dataloader):
inputs = inputs.cuda(0)
logits = prompt_model(inputs)
probs = torch.nn.functional.softmax(logits, dim=-1)
labels = inputs['label']
alllabels.extend(labels.cpu().tolist())
allprobs.extend([max(prob.cpu().tolist()) for prob in probs])
allpreds.extend(torch.argmax(logits, dim=-1).cpu().tolist())
os.makedirs(f"./results/scale/{dataset_name}/{model_name}/{scale}", exist_ok=True)
np.save(f"./results/scale/{dataset_name}/{model_name}/{scale}/alllabels.npy", alllabels)
np.save(f"./results/scale/{dataset_name}/{model_name}/{scale}/allprobs.npy", allprobs)
np.save(f"./results/scale/{dataset_name}/{model_name}/{scale}/allpreds.npy", allpreds)
acc = sum([int(i==j) for i,j in zip(allpreds, alllabels)])/len(allpreds)
print('acc:', acc)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument('--repeats', type=int, default=1)
parser.add_argument('--model_name', type=str, default="t5")
parser.add_argument('--dataset_name', type=str, default="sst2")
parser.add_argument('--scale', type=str, default="base")
args = parser.parse_args()
if args.dataset_name == "yahoo":
args.dataset_name = "yahoo_answers_topics"
args.model_path = MODEL_PATH[f"{args.model_name}-{args.scale}"]
print(args.model_path)
args.dataset_path = DATASET_PATH[args.dataset_name]
args.num_classes = NUM_CLASSES[args.dataset_name]
for i in range(args.repeats):
set_seed(i)
args.seed = i
main(args)