Text-to-Image
MiniMax H3
Diffusers
Safetensors
English
image-to-image
image-editing
image-generation
Instructions to use MiniMaxAI/MiniMax-H3-Image with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use MiniMaxAI/MiniMax-H3-Image with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("MiniMaxAI/MiniMax-H3-Image", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
File size: 1,788 Bytes
7b6142f | 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 | #!/usr/bin/env python3
"""Send request.t2i.json to a MiniMax-H3-Image server and write the PNG.
python3 t2i.py [--url http://127.0.0.1:30000] [--prompt FILE] [--size WxH] [--out t2i.png]
--prompt and --size replace the prompt and the size in the request.
Uses only the Python standard library.
"""
import argparse
import base64
import json
import urllib.request
from pathlib import Path
# absolute(), not resolve(): in the Hugging Face cache this file is a symlink into blobs/.
HERE = Path(__file__).absolute().parent
def main() -> None:
parser = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
parser.add_argument("--url", default="http://127.0.0.1:30000", help="server base URL")
parser.add_argument("--request", type=Path, default=HERE / "request.t2i.json", help="JSON request body")
parser.add_argument("--prompt", type=Path, help="prompt file; replaces the prompt in the request")
parser.add_argument("--size", help="WIDTHxHEIGHT; replaces the size in the request")
parser.add_argument("--out", default="t2i.png")
args = parser.parse_args()
body = json.loads(args.request.read_text())
if args.prompt:
body["prompt"] = args.prompt.read_text()
if args.size:
body["size"] = args.size
request = urllib.request.Request(
f"{args.url.rstrip('/')}/v1/images/generations",
data=json.dumps(body).encode(),
headers={"Content-Type": "application/json"},
)
with urllib.request.urlopen(request, timeout=1800) as response:
payload = json.load(response)
with open(args.out, "wb") as image:
image.write(base64.b64decode(payload["data"][0]["b64_json"]))
print(f"wrote {args.out}")
if __name__ == "__main__":
main()
|