Metadata-Version: 2.4
Name: omi-med-stt
Version: 0.1.0
Summary: CLI for Omi Med STT v1 medical speech-to-text
Author: Omi Health
License: CC-BY-4.0
Requires-Python: >=3.10
Requires-Dist: huggingface-hub<1.0,>=0.23
Requires-Dist: numpy<2.3,>=1.24
Requires-Dist: soundfile>=0.12
Provides-Extra: dev
Requires-Dist: pytest>=8; extra == 'dev'
Provides-Extra: mlx
Requires-Dist: mlx-audio; extra == 'mlx'
Requires-Dist: parakeet-mlx; extra == 'mlx'
Provides-Extra: nemo
Requires-Dist: nemo-toolkit[asr]; extra == 'nemo'
Description-Content-Type: text/markdown

# omi-med-stt

Command-line interface for Omi Med STT v1.

## Install

Until PyPI publication, download the private Hugging Face wheel with `huggingface_hub` and install it locally:

```bash
python - <<'PY'
from huggingface_hub import hf_hub_download
print(hf_hub_download(
    "omi-health/omi-med-stt-v1",
    "packages/omi-med-stt/dist/omi_med_stt-0.1.0-py3-none-any.whl",
))
PY
pip install /path/printed/above/omi_med_stt-0.1.0-py3-none-any.whl
```

After PyPI publication:

```bash
pip install omi-med-stt
```

For Apple Silicon / MLX:

```bash
pip install "omi-med-stt[mlx]"
```

For CUDA/Linux NeMo:

```bash
pip install "omi-med-stt[nemo]"
```

## Usage

Simple path:

```bash
omi-med-stt audio.wav
```

On Apple Silicon this defaults to the MLX repo:

```bash
omi-med-stt audio.wav --runtime mlx --model omi-health/omi-med-stt-v1-mlx
```

On Linux/CUDA this defaults to the NeMo repo:

```bash
omi-med-stt audio.wav --runtime nemo --model omi-health/omi-med-stt-v1
```

JSON output:

```bash
omi-med-stt audio.wav --json
```

Dependency check:

```bash
omi-med-stt check
```

`omi-stt` remains available as a backwards-compatible command alias.

## Long Audio

Omi Med STT v1 is based on Parakeet and can handle long-form audio. You do not need to chunk normal long consultations just to use the model.

An explicit chunked path is still available for constrained environments:

```bash
omi-med-stt transcribe-long consult.wav --chunk-seconds 25 --overlap 3
```

Use the simple `omi-med-stt audio.wav` path first.
