mirror of https://github.com/sean1832/GPT-Brain
You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
59 lines
1.9 KiB
Python
59 lines
1.9 KiB
Python
import openai
|
|
import textwrap
|
|
|
|
from modules import utilities as util
|
|
from modules import gpt_util as gpt
|
|
|
|
|
|
openai.api_key = util.read_file(r'.user\API-KEYS.txt').strip()
|
|
|
|
prompt_dir = '.user/prompt'
|
|
|
|
|
|
def build(chunk_size=4000):
|
|
all_text = util.read_file(r'.user\input.txt')
|
|
|
|
# split text into smaller chunk of 4000 char each
|
|
chunks = textwrap.wrap(all_text, chunk_size)
|
|
|
|
result = []
|
|
|
|
print('Building brain data...')
|
|
for chunk in chunks:
|
|
embedding = gpt.embedding(chunk.encode(encoding='ASCII', errors='ignore').decode())
|
|
info = {'content': chunk, 'vector': embedding}
|
|
print(info, '\n\n\n')
|
|
result.append(info)
|
|
|
|
util.write_json_file(result, r'.user\brain-data.json')
|
|
|
|
|
|
def run_answer(query, model, temp, max_tokens, top_p, freq_penl, pres_penl, chunk_count):
|
|
brain_data = util.read_json_file(r'.user\brain-data.json')
|
|
results = gpt.search_chunks(query, brain_data, chunk_count)
|
|
answers = []
|
|
for result in results:
|
|
my_info = util.read_file(f'{prompt_dir}/my-info.txt')
|
|
|
|
prompt = util.read_file(f'{prompt_dir}/question.txt')
|
|
prompt = prompt.replace('<<INFO>>', result['content'])
|
|
prompt = prompt.replace('<<QS>>', query)
|
|
prompt = prompt.replace('<<MY-INFO>>', my_info)
|
|
|
|
answer = gpt.gpt3(prompt, model, temp, max_tokens, top_p, freq_penl, pres_penl)
|
|
answers.append(answer)
|
|
|
|
all_answers = '\n\n'.join(answers)
|
|
return all_answers
|
|
|
|
|
|
def run(query, model, prompt_file, temp, max_tokens, top_p, freq_penl, pres_penl):
|
|
chunks = textwrap.wrap(query, 10000)
|
|
responses = []
|
|
for chunk in chunks:
|
|
prompt = util.read_file(prompt_file).replace('<<DATA>>', chunk)
|
|
response = gpt.gpt3(prompt, model, temp, max_tokens, top_p, freq_penl, pres_penl)
|
|
responses.append(response)
|
|
all_response = '\n\n'.join(responses)
|
|
return all_response
|