llama-index

1
import os
2
import sys
3
import time
4

5
import openai
6
from openai import OpenAI
7
from validate_json import validate_json
8

9
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
10

11

12
def launch_training(data_path: str) -> None:
13
    validate_json(data_path)
14

15
    # TODO: figure out how to specify file name in the new API
16
    # file_name = os.path.basename(data_path)
17

18
    # upload file
19
    with open(data_path, "rb") as f:
20
        output = client.files.create(
21
            file=f,
22
            purpose="fine-tune",
23
        )
24
    print("File uploaded...")
25

26
    # launch training
27
    while True:
28
        try:
29
            client.fine_tunes.create(training_file=output.id, model="gpt-3.5-turbo")
30
            break
31
        except openai.BadRequestError:
32
            print("Waiting for file to be ready...")
33
            time.sleep(60)
34
    print(f"Training job {output.id} launched. You will be emailed when it's complete.")
35

36

37
if __name__ == "__main__":
38
    data_path = sys.argv[1]
39
    if not os.path.exists(data_path):
40
        raise ValueError(f"Path {data_path} does not exist")
41
    launch_training(data_path)
42