Image-Text-to-Text
Transformers
ONNX
Safetensors
English
medical
chest-xray
radiology
clip
blip
multimodal
cpu
Instructions to use GAD-Research-Lab/MedicalAI-Light-Weight with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use GAD-Research-Lab/MedicalAI-Light-Weight with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="GAD-Research-Lab/MedicalAI-Light-Weight")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("GAD-Research-Lab/MedicalAI-Light-Weight", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use GAD-Research-Lab/MedicalAI-Light-Weight with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "GAD-Research-Lab/MedicalAI-Light-Weight" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GAD-Research-Lab/MedicalAI-Light-Weight", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/GAD-Research-Lab/MedicalAI-Light-Weight
- SGLang
How to use GAD-Research-Lab/MedicalAI-Light-Weight with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "GAD-Research-Lab/MedicalAI-Light-Weight" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GAD-Research-Lab/MedicalAI-Light-Weight", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "GAD-Research-Lab/MedicalAI-Light-Weight" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GAD-Research-Lab/MedicalAI-Light-Weight", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use GAD-Research-Lab/MedicalAI-Light-Weight with Docker Model Runner:
docker model run hf.co/GAD-Research-Lab/MedicalAI-Light-Weight
File size: 7,987 Bytes
e93bfbd | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 | """
Update script — pulls the newest models, UI files, and dependencies.
Usage:
python update.py # interactive menu
python update.py --all # update everything
python update.py --models # download latest ONNX models
python update.py --code # git pull latest source
python update.py --deps # upgrade pip packages
The default model source is a Hugging Face repo.
Configure with: python update.py --set-url <HF_REPO_ID>
"""
import argparse
import json
import os
import subprocess
import sys
from pathlib import Path
CONFIG_FILE = "update_config.json"
DEFAULT_HF_REPO = "your-org/medicalai-models"
DEFAULT_DIR = Path("models") / "default"
ONNX_FULL_DIR = Path("checkpoints") / "onnx_full"
def load_config():
if os.path.exists(CONFIG_FILE):
with open(CONFIG_FILE) as f:
return json.load(f)
return {"hf_repo": DEFAULT_HF_REPO, "auto_update": True}
def save_config(cfg):
with open(CONFIG_FILE, "w") as f:
json.dump(cfg, f, indent=2)
def _ensure_dir(path):
os.makedirs(path, exist_ok=True)
def _file_size(path):
return os.path.getsize(path) / 1024 / 1024
def update_models_from_hf():
"""Download latest ONNX models from Hugging Face."""
cfg = load_config()
repo = cfg["hf_repo"]
from rich.console import Console
console = Console()
if repo == DEFAULT_HF_REPO and "your-org" in repo:
console.print("[yellow]HF_REPO not set. Skipping model download.[/yellow]")
console.print("[yellow]Set your model repo: python update.py --set-url your-org/your-repo[/yellow]")
console.print("[yellow]Or train locally: python training.py --mode prepare-data && python training.py --mode train[/yellow]")
return
try:
import requests
except ImportError:
console.print("[red]'requests' required. Install: pip install requests[/red]")
return
_ensure_dir(ONNX_FULL_DIR)
_ensure_dir(DEFAULT_DIR)
files_to_download = [
("fusion_full.onnx", ONNX_FULL_DIR / "fusion_full.onnx"),
("labels.json", ONNX_FULL_DIR / "labels.json"),
("fusion_classifier.onnx", DEFAULT_DIR / "fusion_classifier.onnx"),
]
base_url = f"https://huggingface.co/{repo}/resolve/main"
for fname, dest in files_to_download:
url = f"{base_url}/{fname}"
console.print(f"[cyan]Downloading {fname}...[/cyan]")
try:
resp = requests.get(url, stream=True, timeout=30)
resp.raise_for_status()
with open(dest, "wb") as f:
for chunk in resp.iter_content(8192):
f.write(chunk)
console.print(f" [green]Saved {dest} ({_file_size(dest):.1f} MB)[/green]")
except Exception as e:
console.print(f" [red]Failed: {e}[/red]")
console.print("[green]Model update complete.[/green]")
def update_code():
"""Pull latest source code from git."""
from rich.console import Console
console = Console()
if not os.path.exists(".git"):
console.print("[yellow]Not a git repository. Skipping code update.[/yellow]")
return
try:
result = subprocess.run(
["git", "pull", "--ff-only"],
capture_output=True, text=True, timeout=60,
)
if result.returncode == 0:
console.print(f"[green]{result.stdout}[/green]")
else:
console.print(f"[yellow]{result.stderr}[/yellow]")
except Exception as e:
console.print(f"[red]Git pull failed: {e}[/red]")
def update_deps():
"""Upgrade all pip packages to latest compatible versions."""
from rich.console import Console
console = Console()
req = "requirements.txt"
if not os.path.exists(req):
console.print("[yellow]No requirements.txt found.[/yellow]")
return
console.print("[cyan]Upgrading dependencies...[/cyan]")
try:
subprocess.check_call(
[sys.executable, "-m", "pip", "install", "--upgrade", "-r", req],
stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
)
console.print("[green]Dependencies upgraded.[/green]")
except Exception as e:
console.print(f"[red]Upgrade failed: {e}[/red]")
def update_all():
from rich.console import Console
console = Console()
console.print("[bold cyan]Full Update[/bold cyan]")
console.print()
console.print("[cyan]Step 1: Updating code...[/cyan]")
update_code()
console.print("[cyan]Step 2: Upgrading dependencies...[/cyan]")
update_deps()
console.print("[cyan]Step 3: Downloading latest models...[/cyan]")
update_models_from_hf()
console.print()
console.print("[green]Update complete![/green]")
console.print(" Run [bold]python quantization.py --mode status[/bold] to verify.")
def set_repo_url(url):
cfg = load_config()
cfg["hf_repo"] = url
save_config(cfg)
print(f"Hugging Face repo set to: {url}")
def show_status():
cfg = load_config()
print(f"Update config: {CONFIG_FILE}")
print(f" HF repo: {cfg['hf_repo']}")
print(f" Auto update: {cfg['auto_update']}")
print()
print("Default models:")
for f in ["fusion_classifier.onnx", "labels.json"]:
p = DEFAULT_DIR / f
exists = os.path.exists(p)
size = f"({_file_size(p):.1f} MB)" if exists else ""
print(f" {f}: {'yes' if exists else 'no'} {size}")
print()
print("Full ONNX pipeline:")
for f in ["fusion_full.onnx", "labels.json"]:
p = ONNX_FULL_DIR / f
exists = os.path.exists(p)
size = f"({_file_size(p):.1f} MB)" if exists else ""
print(f" {f}: {'yes' if exists else 'no'} {size}")
print()
def main():
parser = argparse.ArgumentParser(description="Update MedicalAI models, code, and deps")
parser.add_argument("--models", action="store_true", help="Download latest ONNX models")
parser.add_argument("--code", action="store_true", help="Git pull latest source")
parser.add_argument("--deps", action="store_true", help="Upgrade pip packages")
parser.add_argument("--all", action="store_true", help="Update everything")
parser.add_argument("--set-url", metavar="HF_REPO", help="Set Hugging Face model repo")
parser.add_argument("--status", action="store_true", help="Show update status")
args = parser.parse_args()
if args.set_url:
set_repo_url(args.set_url)
return
if args.status:
show_status()
return
if args.all:
update_all()
return
if args.models:
update_models_from_hf()
return
if args.code:
update_code()
return
if args.deps:
update_deps()
return
# Interactive mode
from rich.console import Console
import questionary
console = Console()
console.print("[bold cyan]MedicalAI - Update Manager[/bold cyan]")
console.print()
choice = questionary.select(
"What would you like to update?",
choices=[
"Everything (code + deps + models)",
"Models only (download latest ONNX)",
"Code only (git pull)",
"Dependencies only (pip upgrade)",
"Show update status",
"Set Hugging Face model repo",
"Cancel",
],
).ask()
if choice == "Everything (code + deps + models)":
update_all()
elif choice == "Models only (download latest ONNX)":
update_models_from_hf()
elif choice == "Code only (git pull)":
update_code()
elif choice == "Dependencies only (pip upgrade)":
update_deps()
elif choice == "Show update status":
show_status()
elif "Set Hugging Face" in choice:
repo = questionary.text("Enter Hugging Face repo (user/repo):").ask()
if repo:
set_repo_url(repo)
else:
console.print("[yellow]Cancelled.[/yellow]")
if __name__ == "__main__":
main()
|