-
Notifications
You must be signed in to change notification settings - Fork 440
/
run_storm_wiki_gpt.py
125 lines (111 loc) · 6.5 KB
/
run_storm_wiki_gpt.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
"""
STORM Wiki pipeline powered by GPT-3.5/4 and You.com search engine.
You need to set up the following environment variables to run this script:
- OPENAI_API_KEY: OpenAI API key
- OPENAI_API_TYPE: OpenAI API type (e.g., 'openai' or 'azure')
- AZURE_API_BASE: Azure API base URL if using Azure API
- AZURE_API_VERSION: Azure API version if using Azure API
- YDC_API_KEY: You.com API key; or, BING_SEARCH_API_KEY: Bing Search API key
Output will be structured as below
args.output_dir/
topic_name/ # topic_name will follow convention of underscore-connected topic name w/o space and slash
conversation_log.json # Log of information-seeking conversation
raw_search_results.json # Raw search results from search engine
direct_gen_outline.txt # Outline directly generated with LLM's parametric knowledge
storm_gen_outline.txt # Outline refined with collected information
url_to_info.json # Sources that are used in the final article
storm_gen_article.txt # Final article generated
storm_gen_article_polished.txt # Polished final article (if args.do_polish_article is True)
"""
import os
import sys
from argparse import ArgumentParser
sys.path.append('./src')
from lm import OpenAIModel
from rm import YouRM, BingSearch
from storm_wiki.engine import STORMWikiRunnerArguments, STORMWikiRunner, STORMWikiLMConfigs
from utils import load_api_key
def main(args):
load_api_key(toml_file_path='secrets.toml')
lm_configs = STORMWikiLMConfigs()
openai_kwargs = {
'api_key': os.getenv("OPENAI_API_KEY"),
'api_provider': os.getenv('OPENAI_API_TYPE'),
'temperature': 1.0,
'top_p': 0.9,
'api_base': os.getenv('AZURE_API_BASE'),
'api_version': os.getenv('AZURE_API_VERSION'),
}
# STORM is a LM system so different components can be powered by different models.
# For a good balance between cost and quality, you can choose a cheaper/faster model for conv_simulator_lm
# which is used to split queries, synthesize answers in the conversation. We recommend using stronger models
# for outline_gen_lm which is responsible for organizing the collected information, and article_gen_lm
# which is responsible for generating sections with citations.
conv_simulator_lm = OpenAIModel(model='gpt-3.5-turbo', max_tokens=500, **openai_kwargs)
question_asker_lm = OpenAIModel(model='gpt-3.5-turbo', max_tokens=500, **openai_kwargs)
outline_gen_lm = OpenAIModel(model='gpt-4-0125-preview', max_tokens=400, **openai_kwargs)
article_gen_lm = OpenAIModel(model='gpt-4-0125-preview', max_tokens=700, **openai_kwargs)
article_polish_lm = OpenAIModel(model='gpt-4-0125-preview', max_tokens=4000, **openai_kwargs)
lm_configs.set_conv_simulator_lm(conv_simulator_lm)
lm_configs.set_question_asker_lm(question_asker_lm)
lm_configs.set_outline_gen_lm(outline_gen_lm)
lm_configs.set_article_gen_lm(article_gen_lm)
lm_configs.set_article_polish_lm(article_polish_lm)
engine_args = STORMWikiRunnerArguments(
output_dir=args.output_dir,
max_conv_turn=args.max_conv_turn,
max_perspective=args.max_perspective,
search_top_k=args.search_top_k,
max_thread_num=args.max_thread_num,
)
# STORM is a knowledge curation system which consumes information from the retrieval module.
# Currently, the information source is the Internet and we use search engine API as the retrieval module.
if args.retriever == 'bing':
rm = BingSearch(bing_search_api=os.getenv('BING_SEARCH_API_KEY'), k=engine_args.search_top_k)
elif args.retriever == 'you':
rm = YouRM(ydc_api_key=os.getenv('YDC_API_KEY'), k=engine_args.search_top_k)
runner = STORMWikiRunner(engine_args, lm_configs, rm)
topic = input('Topic: ')
runner.run(
topic=topic,
do_research=args.do_research,
do_generate_outline=args.do_generate_outline,
do_generate_article=args.do_generate_article,
do_polish_article=args.do_polish_article,
)
runner.post_run()
runner.summary()
if __name__ == '__main__':
parser = ArgumentParser()
# global arguments
parser.add_argument('--output-dir', type=str, default='./results/gpt',
help='Directory to store the outputs.')
parser.add_argument('--max-thread-num', type=int, default=3,
help='Maximum number of threads to use. The information seeking part and the article generation'
'part can speed up by using multiple threads. Consider reducing it if keep getting '
'"Exceed rate limit" error when calling LM API.')
parser.add_argument('--retriever', type=str, choices=['bing', 'you'],
help='The search engine API to use for retrieving information.')
# stage of the pipeline
parser.add_argument('--do-research', action='store_true',
help='If True, simulate conversation to research the topic; otherwise, load the results.')
parser.add_argument('--do-generate-outline', action='store_true',
help='If True, generate an outline for the topic; otherwise, load the results.')
parser.add_argument('--do-generate-article', action='store_true',
help='If True, generate an article for the topic; otherwise, load the results.')
parser.add_argument('--do-polish-article', action='store_true',
help='If True, polish the article by adding a summarization section and (optionally) removing '
'duplicate content.')
# hyperparameters for the pre-writing stage
parser.add_argument('--max-conv-turn', type=int, default=3,
help='Maximum number of questions in conversational question asking.')
parser.add_argument('--max-perspective', type=int, default=3,
help='Maximum number of perspectives to consider in perspective-guided question asking.')
parser.add_argument('--search-top-k', type=int, default=3,
help='Top k search results to consider for each search query.')
# hyperparameters for the writing stage
parser.add_argument('--retrieve-top-k', type=int, default=3,
help='Top k collected references for each section title.')
parser.add_argument('--remove-duplicate', action='store_true',
help='If True, remove duplicate content from the article.')
main(parser.parse_args())