File size: 2,281 Bytes
f346832 | 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 | """Small text-only inference example for the self-contained FxAI checkpoint."""
from __future__ import annotations
import argparse
import json
import sys
from pathlib import Path
# Running the example inside an upload folder should not create *.pyc files.
sys.dont_write_bytecode = True
from futurix_decision import DecisionInput, load_model
def parse_option(value: str) -> tuple[str, str]:
name, separator, description = value.partition("=")
if not separator or not name.strip() or not description.strip():
raise argparse.ArgumentTypeError("options must use NAME=DESCRIPTION")
return name.strip(), description.strip()
def main() -> None:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--checkpoint", type=Path, default=Path(__file__).resolve().parent)
parser.add_argument("--device", default="cuda:0")
parser.add_argument(
"--state",
default="My integration keeps failing. Please help.",
help="Text to classify.",
)
parser.add_argument(
"--instructions",
default="Which team should handle this request?",
help="Question shown to the decision model.",
)
parser.add_argument(
"--option",
action="append",
type=parse_option,
metavar="NAME=DESCRIPTION",
help="Candidate decision option; repeat for multiple options.",
)
parser.add_argument("--max-image-pixels", type=int, default=512 * 512)
args = parser.parse_args()
options = args.option or [
("billing", "Charges and refunds"),
("support", "Technical integration errors"),
("sales", "Questions about purchasing the product"),
]
if len({name for name, _ in options}) != len(options):
parser.error("option names must be unique")
model = load_model(
args.checkpoint,
family="fxai",
device=args.device,
max_image_pixels=args.max_image_pixels,
)
row = DecisionInput(
state=args.state,
question=args.instructions,
options=[name for name, _ in options],
candidate_descriptions=[description for _, description in options],
)
print(json.dumps(model.decide(row).to_dict(), indent=2))
if __name__ == "__main__":
main()
|