### Install and Run OpenMed Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/index.md

Shows how to install the OpenMed library with Hugging Face support using `uv` and then run a Python example script. This is a common workflow for setting up and testing OpenMed functionalities.

```bash
uv pip install "openmed[hf]"
uv run python examples/pii_model_comparison.py
```

--------------------------------

### Install uv and OpenMed with Extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Install uv for dependency management and then install OpenMed with specific extras like Hugging Face support and documentation tooling.

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh  # install uv (skip if already installed)
uv venv --python 3.11                           # create a dedicated virtualenv
source .venv/bin/activate                       # or use `uv python` directly

# install OpenMed with Hugging Face extras and doc tooling
uv pip install ".[hf]"
```

--------------------------------

### Setup OpenMed Environment

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/custom_tokenizer/README.md

Installs necessary dependencies and the OpenMed package in a virtual environment.

```bash
python -m venv .venv-openmed
source .venv-openmed/bin/activate
pip install -r requirements.txt
pip install -e .   # from repo root to get openmed package
```

--------------------------------

### Set Up Development Environment

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/onboarding.md

Clone the repository, create a virtual environment, and install development dependencies. Optionally, install pre-commit hooks.

```bash
# Fork and clone the repo
git clone https://github.com/<your-username>/openmed.git
cd openmed

# Create the local virtual environment and install dev dependencies
uv venv
uv pip install -e ".[dev]"

# (Optional) Install pre-commit hooks
pre-commit install
```

--------------------------------

### Install Service Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

Installs the necessary dependencies for the service using uv pip. Ensure you are in the project root directory.

```bash
uv pip install -e ".[hf,service]"
```

--------------------------------

### Install and Run Privacy Filter Studio

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/privacy_filter_studio/README.md

Install the necessary packages and run the Uvicorn server to launch the Privacy Filter Studio. Ensure you use the correct installation command based on your environment (Apple Silicon vs. other hosts).

```bash
pip install -e " B.[mlx,service]"      # or ".[hf,service]" off Apple Silicon
python -m uvicorn examples.privacy_filter_studio.app:app --reload --port 8770
```

--------------------------------

### Install MLX Backend and Check Backend Type

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/examples.md

Installs the MLX backend and verifies the backend type using a Python one-liner. Ensure you have the necessary dependencies installed.

```bash
uv pip install "".[mlx]"
uv run python -c "from openmed.core.backends import get_backend; print(type(get_backend()).__name__)"
```

--------------------------------

### Install OpenMED Service

Source: https://github.com/maziyarpanahi/openmed/blob/master/README.md

Installs the necessary packages for the OpenMED service and starts the uvicorn server. Ensure you have Docker installed for the Docker command.

```bash
pip install "openmed[hf,service]"
uvicorn openmed.service.app:app --host 0.0.0.0 --port 8080

# or with Docker
docker build -t openmed:1.6.0 .
docker run --rm -p 8080:8080 -e OPENMED_PROFILE=prod openmed:1.6.0
```

--------------------------------

### Install Documentation Dependencies and Serve Locally

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/contributing.md

Install dependencies required for documentation building and serve the MkDocs preview locally with hot reloading. Also includes commands to inspect the built site.

```bash
uv pip install ".[docs]"
uv run mkdocs serve -a 127.0.0.1:8008
make docs-stage
python3 -m http.server --directory site 9000
```

--------------------------------

### Install OpenMed and Testing Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/tests/README.md

Install the package in editable mode with testing dependencies. Ensure transformers and optionally torch are installed.

```bash
# from the repository root
git clone https://github.com/maziyarpanahi/openmed.git
cd openmed
python -m venv .venv
source .venv/bin/activate  # Windows: .venv\Scripts\activate
pip install -e .
pip install pytest pytest-cov transformers
# Optional: install a CPU backend for transformers
pip install torch --index-url https://download.pytorch.org/whl/cpu
```

--------------------------------

### Start API Server with Development Profile

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

Starts the API server using the 'dev' profile. This profile might enable additional debugging or development-specific features.

```bash
OPENMED_PROFILE=dev uvicorn openmed.service.app:app --host 0.0.0.0 --port 8080
```

--------------------------------

### Install Dependencies and Run Linters

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/testing.md

Installs development and Hugging Face dependencies, then runs linting and formatting checks. Use this command to ensure code quality and style compliance.

```bash
uv pip install "[dev,hf]"
make lint
make format-check
make lint-swift             # for Swift/OpenMedKit changes
```

--------------------------------

### Install OpenMed

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/website/index.html

Install the OpenMed library using pip. Use the `mlx` extra for MLX acceleration.

```bash
pip install openmed[mlx]
```

--------------------------------

### Install OpenMed with REST Service

Source: https://github.com/maziyarpanahi/openmed/blob/master/README.md

Install OpenMed with the Hugging Face runtime and the REST service for API access.

```bash
pip install "openmed[hf,service]"
```

--------------------------------

### Install Development Dependencies and Format Code

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/contributing.md

Install the necessary development dependencies and apply canonical code formatting and linting using Ruff. This ensures consistency with CI checks.

```bash
uv pip install -e ".[dev]"
make format
make lint
make format-check
```

--------------------------------

### Quick Start: Text Generation with MLX-LM

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/mlx-backend.md

Generate text using MLX-LM causal language models with OpenMed. This example uses the 'OpenMed/laneformer-2b-it-q4-mlx' model for text generation.

```python
from openmed import generate_text

response = generate_text(
    messages=[
        {
            "role": "user",
            "content": "Explain why local clinical language models matter.",
        }
    ],
    model_name="OpenMed/laneformer-2b-it-q4-mlx",
    max_tokens=128,
)
print(response)
```

--------------------------------

### Install Combined OpenMed Extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Install Hugging Face, MLX runtime, and documentation extras for a full launch surface on one machine.

```bash
uv pip install ".[hf,mlx,docs]"
```

--------------------------------

### Install OpenMed with spaCy extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/spacy-component.md

Install OpenMed including the necessary spaCy integration.

```bash
pip install "openmed[spacy]"
```

--------------------------------

### Install OpenMed with PaddleOCR Backend

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Install the PaddleOCR backend for a heavier opt-in OCR solution.

```bash
uv pip install ".[ocr-paddle]"
```

--------------------------------

### Install Pre-commit Hooks

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/security/secret-handling.md

Install the repository's pre-commit hooks to enforce local checks before committing code. This includes secret scanning.

```bash
pre-commit install
```

--------------------------------

### Install OpenMed

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/PII_Detection_Complete_Guide.ipynb

Install the OpenMed library using pip. This is a prerequisite for using any of its functionalities.

```bash
pip install openmed
```

--------------------------------

### Start API Server

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

Starts the FastAPI API server locally. This command assumes the 'app' object is correctly defined in 'openmed.service.app'.

```bash
uvicorn openmed.service.app:app --host 0.0.0.0 --port 8080
```

--------------------------------

### Example Custom Profile File (TOML)

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/profiles.md

Illustrates the structure and content of a custom profile stored in a TOML file.

```toml
# OpenMed profile: myproject
# Custom profile configuration

log_level = "INFO"
timeout = 450
device = "cuda"
use_medical_tokenizer = true
```

--------------------------------

### Setup: Import OpenMed API and Select Model

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/Deidentification_Cookbook.ipynb

Imports the necessary OpenMed API components and selects a lightweight English PII model for CPU-friendly runs. This setup is reused across all recipes.

```python
import csv
import tempfile
from pathlib import Path

from openmed import (
    deidentify,
    reidentify,
    BatchProcessor,
    DEFAULT_PII_MODELS,
)

# 44M English PII model - the lightest default, good for CPU-only runs.
MODEL = DEFAULT_PII_MODELS["en"]
print("Using model:", MODEL)

# Shared synthetic notes (fabricated PHI).
NOTES = [
    "Patient John Doe (DOB 01/15/1970) called from 555-123-4567.",
    "Contact Jane Roe at jane.roe@example.com regarding MRN 00123456.",
    "Dr. Alan Grant reviewed the chart for Maria Lopez on 2025-03-02.",
]
```

--------------------------------

### Preload Models at Startup

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

Starts the API server with specified models preloaded. Models are specified as a comma-separated list of registry aliases or Hugging Face IDs.

```bash
OPENMED_SERVICE_PRELOAD_MODELS=disease_detection_superclinical,OpenMed/OpenMed-PII-SuperClinical-Small-44M-v1 \
uvicorn openmed.service.app:app --host 0.0.0.0 --port 8080
```

--------------------------------

### Add New Language Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/privacy_filter_multilingual_studio/README.md

To add a new language or example to the studio, edit the `examples.py` file. Update `LANGUAGE_META` if the language is missing and add `StudioExample` tuples to `LANGUAGE_EXAMPLES` for the specific language code. No JavaScript changes are needed as the frontend automatically picks up these changes.

```python
1. Add the language to `LANGUAGE_META` if missing.
2. Add a tuple of `StudioExample(...)` entries to `LANGUAGE_EXAMPLES[code]`.
```

--------------------------------

### Install Editable Package with Dev Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/AGENTS.md

Installs the package in editable mode along with dependencies required for development and testing.

```bash
uv pip install -e ".[dev]"
```

--------------------------------

### Install OpenMed with Apple Silicon Acceleration

Source: https://github.com/maziyarpanahi/openmed/blob/master/README.md

Install OpenMed with support for Apple Silicon acceleration using the MLX runtime.

```bash
pip install "openmed[mlx]"
```

--------------------------------

### Install GLiNER Optional Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/zero-shot-ner.md

Install the GLiNER optional dependencies using uv pip. Ensure GLiNER checkpoints are available locally for inference.

```bash
uv pip install ".[gliner]"
```

--------------------------------

### Install OpenMed with LangChain extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/integrations-langchain.md

Install the OpenMed package with the necessary extras for LangChain integration.

```bash
pip install "openmed[langchain]"
```

--------------------------------

### Accessing Cold Start Latency

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/eval-harness.md

Demonstrates how to access and print the cold start latency from a benchmark report. This metric is useful for observing model load overhead.

```python
report = run_benchmark(fixtures, suite="my-suite", model_name="my-model", runner=runner)
cold_ms = report.metrics["latency"]["cold_start_ms"]
print(f"Cold-start latency: {cold_ms:.1f} ms")
```

--------------------------------

### Start Jupyter Notebook Server

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/README.md

Launch the Jupyter notebook server from your terminal to access and run the notebooks.

```bash
jupyter notebook
```

--------------------------------

### Install OpenMed for OCR and Tesseract

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Install the multimodal extra for scanned images and document OCR, along with the system Tesseract binary.

```bash
uv pip install ".[multimodal]"
brew install tesseract             # macOS
sudo apt-get install tesseract-ocr # Debian/Ubuntu
```

--------------------------------

### Run Tokenization Comparison Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/medical-tokenizer.md

Run the script to compare tokenization outputs from WordPiece, spaCy, and the medical pre-tokenizer on clinical text.

```bash
.venv-openmed/bin/python examples/custom_tokenizer/eval_tokenization_comparison.py
```

--------------------------------

### Install OpenMed with GLiNER or Dev Extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Add GLiNER and transformers support, or install development tools like pytest and coverage.

```bash
uv pip install ".[hf,gliner]"      # add GLiNER + transformers
uv pip install ".[dev]"            # pytest + coverage + linting
```

--------------------------------

### Show CLI Example for OpenMed

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/Medical_Tokenizer_Demo.ipynb

Illustrates the command-line interface usage for analyzing text with OpenMed, specifically showing how to disable the medical tokenizer.

```bash
# Show CLI example (string, not executed here)
cli_cmd = (
    "openmed analyze --text "
    "COVID-19 patient on IL-6 inhibitor"
    " --no-medical-tokenizer"
)
print("CLI example:", cli_cmd)
```

--------------------------------

### Install Runtime Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/Sentence_Detection_Batching.ipynb

Installs necessary libraries for OpenMed NER, including transformers, torch, and pysbd. Ensure you are in the project's virtual environment before running.

```bash
uv pip install transformers==4.57.1 torch==2.9.0 --index-url https://download.pytorch.org/whl/cpu
uv pip install pysbd==0.3.4
```

--------------------------------

### Install Editable Package with Hugging Face Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/AGENTS.md

Installs the package in editable mode, including Hugging Face dependencies for model-related tasks.

```bash
uv pip install -e ".[dev,hf]"
```

--------------------------------

### Start Docker Compose Services

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

Builds and starts the OpenMED service using Docker Compose. This command assumes a docker-compose.yml file is present in the current directory.

```bash
docker compose up -d
```

--------------------------------

### Run Custom Tokenizer Alignment Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/medical-tokenizer.md

Execute the custom tokenizer alignment script to see how custom tokens are mapped through the model and back.

```bash
.venv-openmed/bin/python examples/custom_tokenizer/custom_tokenize_alignment.py
```

--------------------------------

### Run Privacy Filter App (Allow Downloads)

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/privacy_filter_book/README.md

Starts the privacy filter application, allowing first-run Hugging Face downloads for MLX and CPU models.

```bash
OPENMED_PRIVACY_FILTER_DOWNLOAD=1 uvicorn examples.privacy_filter_book.app:app --reload --port 8765
```

--------------------------------

### Install OpenMed and Jupyter

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/README.md

Use uv pip to install OpenMed with HuggingFace support and Jupyter for running notebooks. Additional packages may be needed for PII-related notebooks.

```bash
uv pip install "openmed[hf]"

uv pip install jupyter

# For PII notebooks, you may need:
uv pip install matplotlib ipython
```

--------------------------------

### Install OpenMed with Hugging Face Runtime

Source: https://github.com/maziyarpanahi/openmed/blob/master/README.md

Install the core OpenMed library along with the Hugging Face runtime for CPU or CUDA support on Linux, macOS, and Windows.

```bash
pip install "openmed[hf]"
```

--------------------------------

### Example Profile Report Output

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/profiling.md

A human-readable format for performance profiling reports, showing total duration, operation count, and a breakdown of individual timings with percentages.

```text
Performance Profile Report
==================================================
Total Duration: 1234.56ms
Operations: 4

Timings:
--------------------------------------------------
  inference.forward                    856.23ms ( 69.4%)
  model_loading                        312.45ms ( 25.3%)
  postprocessing                        45.67ms (  3.7%)
  preprocessing                         20.21ms (  1.6%)
```

--------------------------------

### Run Privacy Filter App (Default)

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/privacy_filter_book/README.md

Starts the privacy filter application with default settings, prioritizing cached local models.

```bash
uvicorn examples.privacy_filter_book.app:app --reload --port 8765
```

--------------------------------

### Sample Automation Pipeline for CI

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/examples.md

A bash script to automate the installation of dependencies, running PII model comparison, building documentation, and performing smoke tests for GLiNER models. Useful for CI/CD pipelines.

```bash
#!/usr/bin/env bash
set -euo pipefail

uv pip install "".[hf,docs]"
python examples/pii_model_comparison.py > artifacts/result.txt
uv run mkdocs build --strict
python scripts/smoke_gliner.py --limit 1 --threshold 0.5
```

--------------------------------

### Get OpenMed Version

Source: https://github.com/maziyarpanahi/openmed/blob/master/SECURITY.md

Use this command to retrieve the installed version of OpenMed. This is useful for reporting vulnerabilities.

```bash
python -c "import openmed; print(openmed.__version__)"
```

--------------------------------

### Serve or Build MkDocs Documentation

Source: https://github.com/maziyarpanahi/openmed/blob/master/AGENTS.md

Commands to either serve the documentation locally for preview or to perform a strict build of the MkDocs documentation.

```bash
make docs-serve / make docs-build
```

--------------------------------

### Install OpenMed with MLX extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/mlx-backend.md

Install OpenMed with the necessary MLX dependencies from the repository root. This command installs MLX, MLX-LM, and related libraries.

```bash
pip install -e ".[mlx]"
```

--------------------------------

### Install OpenMed and Dependencies

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/getting_started.ipynb

Installs the OpenMed library and necessary dependencies like transformers and torch.

```python
# Install OpenMed
%pip install openmed

# Install required dependencies
%pip install transformers torch
```

--------------------------------

### Swift MLX Quick Start

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/mlx-backend.md

Demonstrates how to download an MLX model and initialize OpenMed with the MLX backend in Swift. This is useful for running OpenMed MLX artifacts directly within Apple applications on Apple Silicon.

```swift
import OpenMedKit

let modelDirectory = try await OpenMedModelStore.downloadMLXModel(
    repoID: "OpenMed/OpenMed-PII-ClinicalE5-Small-33M-v1-mlx"
)

let openmed = try OpenMed(
    backend: .mlx(modelDirectoryURL: modelDirectory)
)

let entities = try openmed.analyzeText(
    "Patient John Doe, DOB 1990-05-15, SSN 123-45-6789"
)
```

--------------------------------

### Run Full Test Suite

Source: https://github.com/maziyarpanahi/openmed/blob/master/tests/README.md

Execute all unit and integration tests from the project root directory.

```bash
pytest
```

--------------------------------

### Manual Local Package Build and Publish

Source: https://github.com/maziyarpanahi/openmed/blob/master/scripts/README.md

Manually build the package using 'python3 -m build' and then publish it locally using 'hatch publish'. This is an alternative to the CI process.

```bash
python3 -m build
hatch publish
```

--------------------------------

### Install OpenMed MLX Runtime

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/getting-started.md

Install the Python MLX runtime, tokenizer, and artifact dependencies for Apple Silicon Macs.

```bash
uv pip install ".[mlx]"            # Python MLX runtime + tokenizer/artifact deps
uv run python -c "from openmed.core.backends import get_backend; print(type(get_backend()).__name__)"
```

--------------------------------

### Run Swift Package Tests

Source: https://github.com/maziyarpanahi/openmed/blob/master/AGENTS.md

Navigates to the Swift package directory and executes its tests.

```bash
cd swift/OpenMedKit && swift test
```

--------------------------------

### Install OpenMed with AWQ support

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/export-awq.md

Install OpenMed with the 'awq' extra to include the AutoAWQ dependency. This ensures the quantization functionality is available.

```bash
uv pip install -e ".[awq]"
```

--------------------------------

### Timeout Error Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

An example of the error envelope when a request exceeds the configured timeout. The 'details' field includes the timeout duration.

```json
{
  "error": {
    "code": "timeout",
    "message": "Request exceeded configured timeout of 300 seconds",
    "details": {
      "timeout_seconds": 300
    }
  }
}
```

--------------------------------

### Build Project Wheels and Source Distributions

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/contributing.md

Generate distributable package formats (wheels and sdists) for the project. This is a prerequisite for publishing to PyPI.

```bash
python3 -m build
```

--------------------------------

### Install OpenMed with Optional Extras

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/ZeroShot_NER_Tour.ipynb

Install OpenMed with the GLiNER and Hugging Face extras for zero-shot NER capabilities. This command is run from the shell.

```bash
pip install ".[hf,gliner]"
```

```bash
pip install ".[hf,gliner]"
```

--------------------------------

### Example PII Visualization

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/PII_Detection_Complete_Guide.ipynb

Demonstrates PII detection and visualization on a sample text. It uses `extract_pii` and the `highlight_entities` function. Ensure the model and confidence threshold are appropriately set.

```python
viz_text = """Patient: Dr. Sarah Johnson
DOB: 03/15/1975
SSN: 123-45-6789
Phone: (555) 123-4567
Email: sarah.j@email.com
Address: 456 Oak Ave, Boston, MA 02115"""

viz_result = extract_pii(
    viz_text,
    model_name='openmed/OpenMed-PII-SuperClinical-Large-434M-v1',
    confidence_threshold=0.5,
    use_smart_merging=True
)

print("=" * 80)
print("VISUALIZATION: Highlighted PII Entities")
print("=" * 80)
print("\nHover over highlighted text to see entity type and confidence.\n")

html = highlight_entities(viz_result.text, viz_result.entities)
display(HTML(html))

print("\nLegend:")
print("  🟥 Pink: Names")
print("  🟩 Green: Dates")
print("  🟦 Blue: SSN")
print("  🟨 Yellow: Phone")
print("  🟧 Orange: Email")
print("  🟪 Purple: Address")
```

--------------------------------

### Validation Error Example

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/rest-service.md

An example of the error envelope when a request fails validation. The 'details' field provides specific information about the validation failures.

```json
{
  "error": {
    "code": "validation_error",
    "message": "Request validation failed",
    "details": [
      {
        "field": "body.text",
        "message": "Text must not be blank",
        "type": "value_error"
      }
    ]
  }
}
```

--------------------------------

### Install Missing Package

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/README.md

Use this command to install a missing Python package when encountering an ImportError. Ensure 'package-name' is replaced with the actual name of the required package.

```python
# Solution: Install missing package
!pip install package-name
```

--------------------------------

### Create and Save a Custom Profile (Python)

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/profiles.md

Define custom settings and save them as a new profile using the save_profile function. This allows for reusable custom configurations.

```python
from openmed.core.config import save_profile

settings = {
    "log_level": "INFO",
    "timeout": 450,
    "device": "cuda",
    "use_medical_tokenizer": True,
}

save_profile("myproject", settings)
```

--------------------------------

### Quick Start: PII Detection with MLX

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/mlx-backend.md

Analyze text for PII using the MLX backend. MLX is automatically detected on Apple Silicon, so no explicit configuration is needed for this basic usage.

```python
from openmed import analyze_text
from openmed.core.config import OpenMedConfig

# MLX is auto-detected on Apple Silicon — no config needed
result = analyze_text(
    "Patient John Doe, DOB 1990-05-15, SSN 123-45-6789",
    model_name="pii_detection",
)
print(result.entities)
```

--------------------------------

### Install OpenMed with Hugging Face support

Source: https://github.com/maziyarpanahi/openmed/blob/master/examples/notebooks/Multilingual_PII_Detection_Guide.ipynb

Install the OpenMed library with Hugging Face integration using pip. This is required for using the library's advanced features.

```bash
uv pip install "openmed[hf]"
```

--------------------------------

### Run Full Test Suite

Source: https://github.com/maziyarpanahi/openmed/blob/master/AGENTS.md

Executes the complete test suite for the project to ensure code quality before submitting a Pull Request.

```bash
.venv/bin/python -m pytest tests/ -q
```

--------------------------------

### Run Tests

Source: https://github.com/maziyarpanahi/openmed/blob/master/docs/onboarding.md

Execute the full test suite or a scoped subset to confirm baseline passes before making changes.

```bash
# Run the full test suite
.venv/bin/python -m pytest tests/ -q

# Or run a scoped subset (replace with the relevant test path)
.venv/bin/python -m pytest tests/unit/test_pii.py -q
```