voice2text 0.1.0__tar.gz

voice2text-0.1.0/.gitattributes

@@ -0,0 +1,2 @@
+# SCM syntax highlighting & preventing 3-way merges
+pixi.lock merge=binary linguist-language=YAML linguist-generated=true -diff

voice2text-0.1.0/.gitignore

@@ -0,0 +1,8 @@
+.venv/
+__pycache__/
+*.pyc
+*.egg-info/
+dist/
+# pixi environments
+.pixi/*
+!.pixi/config.toml

voice2text-0.1.0/PKG-INFO

@@ -0,0 +1,56 @@
+Metadata-Version: 2.4
+Name: voice2text
+Version: 0.1.0
+Summary: Local voice-to-text with Whisper + LLM cleanup
+Requires-Python: >=3.11
+Requires-Dist: loguru
+Requires-Dist: mlx-whisper
+Requires-Dist: numpy
+Requires-Dist: pynput
+Requires-Dist: scipy
+Requires-Dist: sounddevice
+Description-Content-Type: text/markdown
+
+# voice2text
+
+Local voice-to-text with Whisper + LLM cleanup. Push-to-talk, pastes at cursor.
+
+> **Note:** Before anyone suggests splitting this into modules and submodules — this is an intentional design choice. I want to demonstrate that in December 2025, you can have a fully local voice-to-text system with automatic cleanup and correction, running almost instantly on consumer hardware, all in ~270 lines of Python.
+
+## Install
+
+### Option 1: Pixi (recommended)
+
+```bash
+pixi run ollama pull qwen2.5:3b
+pixi run v2t
+```
+
+### Option 2: UV
+
+```bash
+brew install ollama
+ollama pull qwen2.5:3b
+uv sync
+uv run v2t
+```
+
+## Usage
+
+```bash
+v2t                      # strict mode (restructures sentences)
+v2t --casual             # light cleanup (punctuation only)
+v2t --pause-music        # pause media while recording (macOS only, requires nowplaying-cli)
+```
+
+Hold **Right Command** to record, release to transcribe and paste.
+
+### `--pause-music` (macOS only)
+
+Pauses any playing media while recording and resumes after. Requires:
+
+```bash
+brew install nowplaying-cli
+```
+
+Not available via pixi/conda-forge.

voice2text-0.1.0/README.md

@@ -0,0 +1,43 @@
+# voice2text
+
+Local voice-to-text with Whisper + LLM cleanup. Push-to-talk, pastes at cursor.
+
+> **Note:** Before anyone suggests splitting this into modules and submodules — this is an intentional design choice. I want to demonstrate that in December 2025, you can have a fully local voice-to-text system with automatic cleanup and correction, running almost instantly on consumer hardware, all in ~270 lines of Python.
+
+## Install
+
+### Option 1: Pixi (recommended)
+
+```bash
+pixi run ollama pull qwen2.5:3b
+pixi run v2t
+```
+
+### Option 2: UV
+
+```bash
+brew install ollama
+ollama pull qwen2.5:3b
+uv sync
+uv run v2t
+```
+
+## Usage
+
+```bash
+v2t                      # strict mode (restructures sentences)
+v2t --casual             # light cleanup (punctuation only)
+v2t --pause-music        # pause media while recording (macOS only, requires nowplaying-cli)
+```
+
+Hold **Right Command** to record, release to transcribe and paste.
+
+### `--pause-music` (macOS only)
+
+Pauses any playing media while recording and resumes after. Requires:
+
+```bash
+brew install nowplaying-cli
+```
+
+Not available via pixi/conda-forge.

voice2text-0.1.0/justfile

@@ -0,0 +1,15 @@
+# Lint with ruff
+lint:
+    uv run ruff check --fix voice2text.py
+
+# Build the package
+build:
+    rm -rf dist/
+    uv run hatch build
+
+# Publish to PyPI
+publish:
+    uv run hatch publish
+
+# Build and publish
+release: build publish