lattifai 1.0.5__py3-none-any.whl → 1.2.0__py3-none-any.whl

{lattifai-1.0.5.dist-info → lattifai-1.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lattifai
-Version: 1.0.5
+Version: 1.2.0
 Summary: Lattifai Python SDK: Seamless Integration with Lattifai's Speech and Video AI Services
 Author-email: Lattifai Technologies <tech@lattifai.com>
 Maintainer-email: Lattice <tech@lattifai.com>
@@ -50,7 +50,8 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Python: <3.15,>=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: lattifai-core>=0.5.1
+Requires-Dist: k2py>=0.2.1
+Requires-Dist: lattifai-core>=0.6.0
 Requires-Dist: lattifai-run>=1.0.1
 Requires-Dist: python-dotenv
 Requires-Dist: lhotse>=1.26.0
@@ -61,11 +62,13 @@ Requires-Dist: tgt
 Requires-Dist: onnx>=1.16.0
 Requires-Dist: onnxruntime
 Requires-Dist: msgpack
+Requires-Dist: scipy!=1.16.3
 Requires-Dist: g2p-phonemizer>=0.4.0
 Requires-Dist: av
 Requires-Dist: wtpsplit>=2.1.7
+Requires-Dist: modelscope==1.33.0
 Requires-Dist: OmniSenseVoice>=0.4.2
-Requires-Dist: nemo_toolkit_asr[asr]>=2.7.0rc3
+Requires-Dist: nemo_toolkit_asr[asr]>=2.7.0rc4
 Requires-Dist: pyannote-audio-notorchdeps>=4.0.2
 Requires-Dist: questionary>=2.0
 Requires-Dist: yt-dlp
@@ -113,11 +116,9 @@ Dynamic: license-file
 
 Advanced forced alignment and subtitle generation powered by [ 🤗 Lattice-1](https://huggingface.co/Lattifai/Lattice-1) model.
 
-> **⚠️ Note on Current Limitations**:
-> 1. **Memory Usage**: We are aware of high memory consumption and are actively working on further optimizations.
-
 ## Table of Contents
 
+- [Core Capabilities](#core-capabilities)
 - [Installation](#installation)
 - [Quick Start](#quick-start)
   - [Command Line Interface](#command-line-interface)
@@ -134,16 +135,45 @@ Advanced forced alignment and subtitle generation powered by [ 🤗 Lattice-1](h
   - [YouTube Processing](#youtube-processing)
   - [Configuration Objects](#configuration-objects)
 - [Advanced Features](#advanced-features)
+  - [Audio Preprocessing](#audio-preprocessing)
+  - [Long-Form Audio Support](#long-form-audio-support)
   - [Word-Level Alignment](#word-level-alignment)
   - [Smart Sentence Splitting](#smart-sentence-splitting)
-  - [Speaker Diarization](#speaker-diarization-wip)
+  - [Speaker Diarization](#speaker-diarization)
   - [YAML Configuration Files](#yaml-configuration-files)
+- [Architecture Overview](#architecture-overview)
+- [Performance & Optimization](#performance--optimization)
 - [Supported Formats](#supported-formats)
+- [Supported Languages](#supported-languages)
 - [Roadmap](#roadmap)
 - [Development](#development)
 
 ---
 
+## Core Capabilities
+
+LattifAI provides comprehensive audio-text alignment powered by the Lattice-1 model:
+
+| Feature | Description | Status |
+|---------|-------------|--------|
+| **Forced Alignment** | Precise word-level and segment-level synchronization with audio | ✅ Production |
+| **Multi-Model Transcription** | Gemini (100+ languages), Parakeet (24 languages), SenseVoice (5 languages) | ✅ Production |
+| **Speaker Diarization** | Automatic multi-speaker identification with label preservation | ✅ Production |
+| **Audio Preprocessing** | Multi-channel selection, device optimization (CPU/CUDA/MPS) | ✅ Production |
+| **Streaming Mode** | Process audio up to 20 hours with minimal memory footprint | ✅ Production |
+| **Smart Text Processing** | Intelligent sentence splitting and non-speech element separation | ✅ Production |
+| **Universal Format Support** | 30+ caption/subtitle formats with text normalization | ✅ Production |
+| **Configuration System** | YAML-based configs for reproducible workflows | ✅ Production |
+
+**Key Highlights:**
+- 🎯 **Accuracy**: State-of-the-art alignment precision with Lattice-1 model
+- 🌍 **Multilingual**: Support for 100+ languages via multiple transcription models
+- 🚀 **Performance**: Hardware-accelerated processing with streaming support
+- 🔧 **Flexible**: CLI, Python SDK, and Web UI interfaces
+- 📦 **Production-Ready**: Battle-tested on diverse audio/video content
+
+---
+
 ## Installation
 
 ### Step 1: Install SDK
@@ -151,9 +181,6 @@ Advanced forced alignment and subtitle generation powered by [ 🤗 Lattice-1](h
 **Using pip:**
 ```bash
 
-pip install install-k2
-install-k2 --torch-version 2.9.1  # if not set will auto-detect PyTorch version and install compatible k2
-
 pip install lattifai
 ```
 
@@ -167,30 +194,11 @@ uv init my-project
 cd my-project
 source .venv/bin/activate
 
-# Install k2 (required dependency)
-uv pip install install-k2
-uv pip install pip
-uv run install-k2 --torch-version 2.9.1
-
 # Install LattifAI
 uv pip install lattifai
 ```
 
-> **Note**: `install-k2` automatically detects your PyTorch version (up to 2.9) and installs the compatible k2 wheel.
 
-<details>
-<summary><b>install-k2 options</b></summary>
-
-```
-usage: install-k2 [-h] [--system {linux,darwin,windows}] [--dry-run] [--torch-version TORCH_VERSION]
-
-optional arguments:
-  -h, help                      Show this help message and exit
-  --system {linux,darwin,windows}  Override OS detection
-  --dry-run                     Show what would be installed without making changes
-  --torch-version TORCH_VERSION    Specify torch version (e.g., 2.8.0)
-```
-</details>
 
 ### Step 2: Get Your API Key
 
@@ -254,7 +262,7 @@ caption = client.alignment(
 
 That's it! Your aligned subtitles are saved to `aligned.srt`.
 
-### Web Interface
+### 🚧 Web Interface
 
 ![web Demo](assets/web.png)
 
@@ -312,13 +320,9 @@ That's it! Your aligned subtitles are saved to `aligned.srt`.
    The web interface will automatically open in your browser at `http://localhost:5173`.
 
 **Features:**
-- ✅ Automatic backend server status detection
-- ✅ Visual file upload with drag-and-drop
-- ✅ Real-time alignment progress
-- ✅ Multiple subtitle format support
-- ✅ Built-in transcription with multiple models
-- ✅ API key management interface
-- ✅ Download aligned subtitles in various formats
+- ✅ **Drag-and-Drop Upload**: Visual file upload for audio/video and captions
+- ✅ **Real-Time Progress**: Live alignment progress with detailed status
+- ✅ **Multiple Transcription Models**: Gemini, Parakeet, SenseVoice selection
 
 ---
 
@@ -619,6 +623,78 @@ from lattifai import (
 
 ## Advanced Features
 
+### Audio Preprocessing
+
+LattifAI provides powerful audio preprocessing capabilities for optimal alignment:
+
+**Channel Selection**
+
+Control which audio channel to process for stereo/multi-channel files:
+
+```python
+from lattifai import LattifAI
+
+client = LattifAI()
+
+# Use left channel only
+caption = client.alignment(
+    input_media="stereo.wav",
+    input_caption="subtitle.srt",
+    channel_selector="left",  # Options: "left", "right", "average", or channel index (0, 1, 2, ...)
+)
+
+# Average all channels (default)
+caption = client.alignment(
+    input_media="stereo.wav",
+    input_caption="subtitle.srt",
+    channel_selector="average",
+)
+```
+
+**CLI Usage:**
+```bash
+# Use right channel
+lai alignment align audio.wav subtitle.srt output.srt \
+    media.channel_selector=right
+
+# Use specific channel index
+lai alignment align audio.wav subtitle.srt output.srt \
+    media.channel_selector=1
+```
+
+**Device Management**
+
+Optimize processing for your hardware:
+
+```python
+from lattifai import LattifAI, AlignmentConfig
+
+# Use CUDA GPU
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="cuda")
+)
+
+# Use specific GPU
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="cuda:0")
+)
+
+# Use Apple Silicon MPS
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="mps")
+)
+
+# Use CPU
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="cpu")
+)
+```
+
+**Supported Formats**
+- **Audio**: WAV, MP3, M4A, AAC, FLAC, OGG, OPUS, AIFF, and more
+- **Video**: MP4, MKV, MOV, WEBM, AVI, and more
+- All formats supported by FFmpeg are compatible
+
 ### Long-Form Audio Support
 
 LattifAI now supports processing long audio files (up to 20 hours) through streaming mode. Enable streaming by setting the `streaming_chunk_secs` parameter:
@@ -660,14 +736,18 @@ client = LattifAI(
 )
 ```
 
-**Notes:**
-- Chunk duration must be between 1 and 1800 seconds (minimum 1 second, maximum 30 minutes)
-- Default value: 600 seconds (10 minutes)
-- **Recommended: Use 60 seconds or larger for optimal performance**
-- Set to `None` to disable streaming
-- **Thanks to our precise implementation, streaming has virtually no impact on alignment accuracy**
-- Smaller chunks reduce memory usage with minimal quality trade-off
-- Recommended chunk size: 300-900 seconds (5-15 minutes) for optimal balance
+**Technical Details:**
+
+| Parameter | Description | Recommendation |
+|-----------|-------------|----------------|
+| **Default Value** | 600 seconds (10 minutes) | Good for most use cases |
+| **Memory Impact** | Lower chunks = less RAM usage | Adjust based on available RAM |
+| **Accuracy Impact** | Virtually zero degradation | Our precise implementation preserves quality |
+
+**Performance Characteristics:**
+- ✅ **Near-Perfect Accuracy**: Streaming implementation maintains alignment precision
+- 🚧 **Memory Efficient**: Process 20-hour audio with <10GB RAM (600-sec chunks)
+
 
 ### Word-Level Alignment
 
@@ -708,18 +788,50 @@ caption = client.alignment(
 )
 ```
 
-### Speaker Diarization (WIP)
+### Speaker Diarization
+
+Speaker diarization automatically identifies and labels different speakers in audio using state-of-the-art models.
+
+**Core Capabilities:**
+- 🎤 **Multi-Speaker Detection**: Automatically detect speaker changes in audio
+- 🏷️ **Smart Labeling**: Assign speaker labels (SPEAKER_00, SPEAKER_01, etc.)
+- 🔄 **Label Preservation**: Maintain existing speaker names from input captions
+- 🤖 **Gemini Integration**: Extract speaker names intelligently during transcription
+
+**How It Works:**
+
+1. **Without Existing Labels**: System assigns generic labels (SPEAKER_00, SPEAKER_01)
+2. **With Existing Labels**: System preserves your speaker names during alignment
+   - Formats: `[Alice]`, `>> Bob:`, `SPEAKER_01:`, `Alice:` are all recognized
+3. **Gemini Transcription**: When using Gemini models, speaker names are extracted from context
+   - Example: "Hi, I'm Alice" → System labels as `Alice` instead of `SPEAKER_00`
+
+**Speaker Label Integration:**
+
+The diarization engine intelligently matches detected speakers with existing labels:
+- If input captions have speaker names → **Preserved during alignment**
+- If Gemini transcription provides names → **Used for labeling**
+- Otherwise → **Generic labels (SPEAKER_00, etc.) assigned**
+    * 🚧 **Future Enhancement:**
+        - **AI-Powered Speaker Name Inference**: Upcoming feature will use large language models combined with metadata (video title, description, context) to intelligently infer speaker names, making transcripts more human-readable and contextually accurate
 
-**Note:** This feature is currently under development and not yet fully available.
+**CLI:**
+```bash
+# Enable speaker diarization during alignment
+lai alignment align audio.wav subtitle.srt output.srt \
+    diarization.enabled=true
 
-Speaker diarization automatically identifies and labels different speakers in audio. When enabled, the system will:
-- Detect speaker changes in the audio
-- Assign speaker labels (e.g., SPEAKER_00, SPEAKER_01) to each segment
-- Update subtitle segments with speaker information
+# With additional diarization settings
+lai alignment align audio.wav subtitle.srt output.srt \
+    diarization.enabled=true \
+    diarization.device=cuda \
+    diarization.min_speakers=2 \
+    diarization.max_speakers=4
 
-**Speaker Name Handling:**
-- **Existing speaker labels in subtitles**: If your input captions already contain speaker names (e.g., `[Alice]`, `>> Bob:`, or `SPEAKER_01:`), the system will preserve them as much as possible during alignment
-- **Gemini Transcriber**: When using Gemini models for transcription (e.g., `gemini-2.5-pro`), the model can intelligently identify and extract speaker names from dialogue context, making it easier to generate speaker-aware transcripts
+# For YouTube videos with diarization
+lai alignment youtube "https://youtube.com/watch?v=VIDEO_ID" \
+    diarization.enabled=true
+```
 
 **Python SDK:**
 ```python
@@ -742,6 +854,8 @@ for segment in caption.supervisions:
 
 ### YAML Configuration Files
 
+* **under development**
+
 Create reusable configuration files:
 
 ```yaml
@@ -758,6 +872,125 @@ lai alignment align audio.wav subtitle.srt output.srt \
 
 ---
 
+## Architecture Overview
+
+LattifAI uses a modular, config-driven architecture for maximum flexibility:
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│                      LattifAI Client                        │
+├─────────────────────────────────────────────────────────────┤
+│  Configuration Layer (Config-Driven)                        │
+│  ├── ClientConfig      (API settings)                       │
+│  ├── AlignmentConfig   (Model & device)                     │
+│  ├── CaptionConfig     (I/O formats)                        │
+│  ├── TranscriptionConfig (ASR models)                       │
+│  └── DiarizationConfig (Speaker detection)                  │
+├─────────────────────────────────────────────────────────────┤
+│  Core Components                                            │
+│  ├── AudioLoader      → Load & preprocess audio             │
+│  ├── Aligner          → Lattice-1 forced alignment          │
+│  ├── Transcriber      → Multi-model ASR                     │
+│  ├── Diarizer         → Speaker identification              │
+│  └── Tokenizer        → Intelligent text segmentation       │
+├─────────────────────────────────────────────────────────────┤
+│  Data Flow                                                  │
+│  Input → AudioLoader → Aligner → Diarizer → Caption        │
+│             ↓                                               │
+│         Transcriber (optional)                              │
+└─────────────────────────────────────────────────────────────┘
+```
+
+**Component Responsibilities:**
+
+| Component | Purpose | Configuration |
+|-----------|---------|---------------|
+| **AudioLoader** | Load audio/video, channel selection, format conversion | `MediaConfig` |
+| **Aligner** | Forced alignment using Lattice-1 model | `AlignmentConfig` |
+| **Transcriber** | ASR with Gemini/Parakeet/SenseVoice | `TranscriptionConfig` |
+| **Diarizer** | Speaker diarization with pyannote.audio | `DiarizationConfig` |
+| **Tokenizer** | Sentence splitting and text normalization | `CaptionConfig` |
+| **Caption** | Unified data structure for alignments | `CaptionConfig` |
+
+**Data Flow:**
+
+1. **Audio Loading**: `AudioLoader` loads media, applies channel selection, converts to numpy array
+2. **Transcription** (optional): `Transcriber` generates transcript if no caption provided
+3. **Text Preprocessing**: `Tokenizer` splits sentences and normalizes text
+4. **Alignment**: `Aligner` uses Lattice-1 to compute word-level timestamps
+5. **Diarization** (optional): `Diarizer` identifies speakers and assigns labels
+6. **Output**: `Caption` object contains all results, exported to desired format
+
+**Configuration Philosophy:**
+- ✅ **Declarative**: Describe what you want, not how to do it
+- ✅ **Composable**: Mix and match configurations
+- ✅ **Reproducible**: Save configs to YAML for consistent results
+- ✅ **Flexible**: Override configs per-method or globally
+
+---
+
+## Performance & Optimization
+
+### Device Selection
+
+Choose the optimal device for your hardware:
+
+```python
+from lattifai import LattifAI, AlignmentConfig
+
+# NVIDIA GPU (recommended for speed)
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="cuda")
+)
+
+# Apple Silicon GPU
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="mps")
+)
+
+# CPU (maximum compatibility)
+client = LattifAI(
+    alignment_config=AlignmentConfig(device="cpu")
+)
+```
+
+**Performance Comparison** (30-minute audio):
+
+| Device | Time |
+|--------|------|
+| CUDA (RTX 4090) | ~18 sec |
+| MPS (M4) | ~26 sec |
+
+### Memory Management
+
+**Streaming Mode** for long audio:
+
+```python
+# Process 20-hour audio with <10GB RAM
+caption = client.alignment(
+    input_media="long_audio.wav",
+    input_caption="subtitle.srt",
+    streaming_chunk_secs=600.0,  # 10-minute chunks
+)
+```
+
+**Memory Usage** (approximate):
+
+| Chunk Size | Peak RAM | Suitable For |
+|------------|----------|-------------|
+| 600 sec | ~5 GB | Recommended |
+| No streaming | ~10 GB+ | Short audio only |
+
+### Optimization Tips
+
+1. **Use GPU when available**: 10x faster than CPU
+2. **WIP: Enable streaming for long audio**: Process 20+ hour files without OOM
+3. **Choose appropriate chunk size**: Balance memory vs. performance
+4. **Batch processing**: Process multiple files in sequence (coming soon)
+5. **Profile alignment**: Set `client.profile=True` to identify bottlenecks
+
+---
+
 ## Supported Formats
 
 LattifAI supports virtually all common media and subtitle formats:
@@ -778,14 +1011,77 @@ LattifAI supports virtually all common media and subtitle formats:
 
 ---
 
+## Supported Languages
+
+LattifAI supports multiple transcription models with different language capabilities:
+
+### Gemini Models (100+ Languages)
+
+**Models**: `gemini-2.5-pro`, `gemini-3-pro-preview`, `gemini-3-flash-preview`
+
+**Supported Languages**: English, Chinese (Mandarin & Cantonese), Spanish, French, German, Italian, Portuguese, Japanese, Korean, Arabic, Russian, Hindi, Bengali, Turkish, Dutch, Polish, Swedish, Danish, Norwegian, Finnish, Greek, Hebrew, Thai, Vietnamese, Indonesian, Malay, Filipino, Ukrainian, Czech, Romanian, Hungarian, Swahili, Tamil, Telugu, Marathi, Gujarati, Kannada, and 70+ more languages.
+
+> **Note**: Requires Gemini API key from [Google AI Studio](https://aistudio.google.com/apikey)
+
+### NVIDIA Parakeet (24 European Languages)
+
+**Model**: `nvidia/parakeet-tdt-0.6b-v3`
+
+**Supported Languages**:
+- **Western Europe**: English (en), French (fr), German (de), Spanish (es), Italian (it), Portuguese (pt), Dutch (nl)
+- **Nordic**: Danish (da), Swedish (sv), Norwegian (no), Finnish (fi)
+- **Eastern Europe**: Polish (pl), Czech (cs), Slovak (sk), Hungarian (hu), Romanian (ro), Bulgarian (bg), Ukrainian (uk), Russian (ru)
+- **Others**: Croatian (hr), Estonian (et), Latvian (lv), Lithuanian (lt), Slovenian (sl), Maltese (mt), Greek (el)
+
+### Alibaba SenseVoice (5 Asian Languages)
+
+**Model**: `iic/SenseVoiceSmall`
+
+**Supported Languages**:
+- Chinese/Mandarin (zh)
+- English (en)
+- Japanese (ja)
+- Korean (ko)
+- Cantonese (yue)
+
+### Language Selection
+
+```python
+from lattifai import LattifAI, TranscriptionConfig
+
+# Specify language for transcription
+client = LattifAI(
+    transcription_config=TranscriptionConfig(
+        model_name="nvidia/parakeet-tdt-0.6b-v3",
+        language="de",  # German
+    )
+)
+```
+
+**CLI Usage:**
+```bash
+lai transcribe run audio.wav output.srt \
+    transcription.model_name=nvidia/parakeet-tdt-0.6b-v3 \
+    transcription.language=de
+```
+
+> **Tip**: Use Gemini models for maximum language coverage, Parakeet for European languages, and SenseVoice for Asian languages.
+
+---
+
 ## Roadmap
 
 Visit our [LattifAI roadmap](https://lattifai.com/roadmap) for the latest updates.
 
-| Date | Release | Features |
+| Date | Model Release | Features |
 |------|---------|----------|
 | **Oct 2025** | **Lattice-1-Alpha** | ✅ English forced alignment<br>✅ Multi-format support<br>✅ CPU/GPU optimization |
-| **Nov 2025** | **Lattice-1** | ✅ English + Chinese + German<br>✅ Mixed languages alignment<br>🚀 Integrate Speaker Diarization |
+| **Nov 2025** | **Lattice-1** | ✅ English + Chinese + German<br>✅ Mixed languages alignment<br>✅ Speaker Diarization<br>✅ Multi-model transcription (Gemini, Parakeet, SenseVoice)<br>✅ Web interface with React<br>🚧 Advanced segmentation strategies (entire/transcription/hybrid)<br>🚧 Audio event detection ([MUSIC], [APPLAUSE], etc.)<br> |
+| **Q1 2026** | **Lattice-2** | ✅ Streaming mode for long audio<br>🔮 40+ languages support<br>🔮 Real-time alignment |
+
+
+
+**Legend**: ✅ Released | 🚧 In Development | 📋 Planned | 🔮 Future
 
 ---
 

{lattifai-1.0.5.dist-info → lattifai-1.2.0.dist-info}/RECORD

@@ -1,44 +1,45 @@
-lattifai/__init__.py,sha256=7y1R5IGw0Sgvl1tfqxEK7e-ozW0wVB-q_JZgv6YyrMQ,2751
-lattifai/audio2.py,sha256=BKMCzkuEmBFAWOEnzgLxeK8TBPTFbjzr1esOfe3MQoo,17460
-lattifai/client.py,sha256=OXDGsWVeOMEjmXI795pvnK3L-ZLn_sfUwG0i7uJ1JkY,22492
+lattifai/__init__.py,sha256=l7dIodSCVMHUXQkd8BVGBoDdYojBCh_lyBWlVibynk8,2695
+lattifai/audio2.py,sha256=P3N8_BwiscbetzDbkbj-n8BcMu2vWD6-MvtQvGwWWf0,17448
+lattifai/client.py,sha256=Vqg4vY--6tox9Js0qGWlE7IGeHJVyQeYLTXYtlzPk3w,19020
 lattifai/errors.py,sha256=LyWRGVhQ6Ak2CYn9FBYAPRgQ7_VHpxzNsXI31HXD--s,11291
 lattifai/logging.py,sha256=MbUEeOUFlF92pA9v532DiPPWKl03S7UHCJ6Z652cf0w,2860
-lattifai/mixin.py,sha256=yj3H1SSQSQrhUeqKhQmRRELRr5fp2mb2ovkK9p8Vwn4,23858
+lattifai/mixin.py,sha256=wdgxEhgxR--dHXmeiJZ4AQDxEjKo49GLYQ0BXJw3qpk,25206
 lattifai/types.py,sha256=SjYBfwrCBOXlICvH04niFQJ7OzTx7oTaa_npfRkB67U,659
-lattifai/utils.py,sha256=ZYEUaoTBCwzv4PBBD-woeiDSTx8T1a1vXHIT0g1YmRI,5345
+lattifai/utils.py,sha256=cMiC5CY6gSMtcOtf_wK1BBMBEfHwc5R_S8_NIoVYk6I,5321
 lattifai/alignment/__init__.py,sha256=ehpkKfjNIYUx7_M-RWD_8Efcrzd9bE-NSm0QgMMVLW0,178
-lattifai/alignment/lattice1_aligner.py,sha256=DpN_it7ETZgz6uH3I90Y926bvjhFRdL6dycxz5S_tkI,5142
-lattifai/alignment/lattice1_worker.py,sha256=1yYK_xLOL_xHZTVGgNb957R7HhHnl6xwrXUcN372ZIY,12407
+lattifai/alignment/lattice1_aligner.py,sha256=wm1BWNu4h-b507OAvLV0ITi7g0qaWthOPwvzWFHKyZQ,6251
+lattifai/alignment/lattice1_worker.py,sha256=ls2o3pVChB63OQrElJOmHzYIhCkjBFPt8EsLIVR1sJ0,11104
 lattifai/alignment/phonemizer.py,sha256=fbhN2DOl39lW4nQWKzyUUTMUabg7v61lB1kj8SKK-Sw,1761
 lattifai/alignment/segmenter.py,sha256=mzWEQC6hWZtI2mR2WU59W7qLHa7KXy7fdU6991kyUuQ,6276
-lattifai/alignment/tokenizer.py,sha256=oqgy5L9wU0_AMyUVNArEtPIDXm7WdvNNfJuB2ZJBpqI,22394
+lattifai/alignment/tokenizer.py,sha256=JY11uEe-v4KQLoHZuaHgdFqgxR3u_1D9ZXXMnB6hA-Q,22994
 lattifai/caption/__init__.py,sha256=6MM_2j6CaqwZ81LfSy4di2EP0ykvheRjMZKAYDx2rQs,477
-lattifai/caption/caption.py,sha256=Ljt-6K89AauIK05hdDqjV6G03mkTTJL2UE9ukt-tck0,52502
+lattifai/caption/caption.py,sha256=mZYobxuZ8tkJUkZMVvRTrNeGTdmIZYSXTEySQdaGQd8,54595
 lattifai/caption/gemini_reader.py,sha256=GqY2w78xGYCMDP5kD5WGS8jK0gntel2SK-EPpPKTrwU,15138
 lattifai/caption/gemini_writer.py,sha256=sYPxYEmVQcEan5WVGgSrcraxs3QJRQRh8CJkl2yUQ1s,6515
 lattifai/caption/supervision.py,sha256=DRrM8lfKU_x9aVBcLG6xnT0xIJrnc8jzHpzcSwQOg8c,905
 lattifai/caption/text_parser.py,sha256=XDb8KTt031uJ1hg6dpbINglGOTX-6pBcghbg3DULM1I,4633
-lattifai/cli/__init__.py,sha256=dIUmrpN-OwR4h6BqMhXp87_5ZwgO41ShPru_iZGnpQs,463
+lattifai/cli/__init__.py,sha256=LafsAf8YfDcfTeJ1IevFcyLm-mNbxpOOnm33OFKtpDM,523
 lattifai/cli/alignment.py,sha256=06em-Uaf6NhSz1ce4dwT2r8n56NrtibR7ZsSkmc18Kc,5954
 lattifai/cli/app_installer.py,sha256=gAndH3Yo97fGRDe2CQnGtOgZZ4k3_v5ftcUo5g6xbSA,5884
-lattifai/cli/caption.py,sha256=p0VY6orf3D77tr30NQka7A84kwEmYiZrCDB6FbTgoFM,6312
+lattifai/cli/caption.py,sha256=4qQ9DFhxcfaeFMY0TB5I42x4W_gOo2zY6kjXnHnFDms,6313
+lattifai/cli/diarization.py,sha256=GTd2vnTm6cJN6Q3mFP-ShY9bZBl1_zKzWFu-4HHcMzk,4075
 lattifai/cli/server.py,sha256=sXMfOSse9-V79slXUU8FDLeqtI5U9zeU-5YpjTIGyVw,1186
-lattifai/cli/transcribe.py,sha256=W42SVhnOQ0EndMk-Lu38BiG1LuMcJnzre9X83M6kBZ4,8137
-lattifai/cli/youtube.py,sha256=-EIDSS1Iel3_6qD9M2CZZHwKOvgdkIa1cMY4rX7xwVo,5331
+lattifai/cli/transcribe.py,sha256=_vHzrdaGiPepQWATqvEDYDjwzfVLAd2i8RjOLkvdb0w,8218
+lattifai/cli/youtube.py,sha256=9M2dpcUCvT7vVbXJCIxJwe9klJXoF2jUeLxiatslYso,6063
 lattifai/config/__init__.py,sha256=Z8OudvS6fgfLNLu_2fvoXartQiYCECOnNfzDt-PfCN4,543
-lattifai/config/alignment.py,sha256=v6SuryAVNET9hgH_ZidYN2QhZqpEDnNhR-rogSSSfAg,4039
+lattifai/config/alignment.py,sha256=vLiH150YWvBUiVkFOIO-nPXCB-b8fP9iSZgS79k1Qbg,4586
 lattifai/config/caption.py,sha256=AYOyUJ1xZsX8CBZy3GpLitbcCAHcZ9LwXui_v3vtuso,6786
-lattifai/config/client.py,sha256=I1JqLQlsQNU5ouovTumr-PP_8GWC9DI_e9B5UwsDZws,1492
+lattifai/config/client.py,sha256=46b816MiYja3Uan_3wjnhtqDr0M6T-FqEygJ3e50IZc,1664
 lattifai/config/diarization.py,sha256=cIkwCfsYqfMns3i6tKWcwBBBkdnhhmB_Eo0TuOPCw9o,2484
 lattifai/config/media.py,sha256=cjM8eGeZ7ELhmy4cCqHAyogeHItaVqMrPzSwwIx79HY,14856
-lattifai/config/transcription.py,sha256=bzghOGgcNWzTnDYd_cqCOB7GT8OnzHDiyam7LSixqxM,2901
+lattifai/config/transcription.py,sha256=_gPJD6cob_jWNdf841nBHhAqJGCxS6PfSyvx2W_vPcM,3082
 lattifai/diarization/__init__.py,sha256=MgBDQ1ehL2qDnZprEp8KqON7CmbG-qaP37gzBsV0jzk,119
-lattifai/diarization/lattifai.py,sha256=SE2BpIZ3_deKyhXdBqe77bsDLXIUV9AQV34gfINv7_s,2657
+lattifai/diarization/lattifai.py,sha256=tCnFL6ywITqeKR8YoCsYvyJxNoIwoC6GsnI9zkXNB-Q,3128
 lattifai/server/app.py,sha256=wXYgXc_yGQACtUJdhkfhLsTOQjhhIhDQRiVRny7Ogcs,15455
-lattifai/transcription/__init__.py,sha256=mEoMTbs5jAgtXQn1jTjlFY_GUr-S0WmPn8uZ6WZCkU0,2643
-lattifai/transcription/base.py,sha256=59b4nQHFMyTRyyzBJTM8ZpEuUy1KjwA2o6rNfrNluKY,3911
-lattifai/transcription/gemini.py,sha256=1VNi9gl-Kpkw3ljZcOZG5oq_OY8fMC9Xv4kOwyQpI0Q,7992
-lattifai/transcription/lattifai.py,sha256=h0nhXST0qljhyndf80IEddM7Y_N1jiS28YoaE536eME,3483
+lattifai/transcription/__init__.py,sha256=vMHciyCEPKhhfM3KjMCeDqnyxU1oghF8g5o5SvpnT_4,2669
+lattifai/transcription/base.py,sha256=v_b1_JGYiBqeMmwns0wHCJ7UOm6j9k-76Uzbr-qmzrs,4467
+lattifai/transcription/gemini.py,sha256=LJSQt9nGqQdEG6ZFXoHWltumyMEM7-Ezy8ss0iPJb7k,12414
+lattifai/transcription/lattifai.py,sha256=EKEdCafgdRWKw_084eD07BqGh2_D-qo3ig3H5X3XYGg,4621
 lattifai/transcription/prompts/README.md,sha256=X49KWSQVdjWxxWUp4R2w3ZqKrAOi6_kDNHh1hMaQ4PE,694
 lattifai/transcription/prompts/__init__.py,sha256=G9b42COaCYv3sPPNkHsGDLOMBuVGKt4mXGYal_BYtYQ,1351
 lattifai/transcription/prompts/gemini/README.md,sha256=rt7f7yDGtaobKBo95LG3u56mqa3ABOXQd0UVgJYtYuo,781
@@ -47,10 +48,10 @@ lattifai/workflow/__init__.py,sha256=GOT9jptXwpIMiNRqJ_LToEt_5Dt0k7XXbLkFzhrl31o
 lattifai/workflow/agents.py,sha256=yEOnxnhcTvr1iOhCorNvp8B76P6nQsLRXJCu_rCYFfM,38
 lattifai/workflow/base.py,sha256=8QoVIBZwJfr5mppJbtUFafHv5QR9lL-XrULjTWD0oBg,6257
 lattifai/workflow/file_manager.py,sha256=IUWW838ta83kfwM4gpW83gsD_Tx-pa-L_RWKjiefQbQ,33017
-lattifai/workflow/youtube.py,sha256=ON9z0UUk16ThQzdhdgyOiwBmewZOcxfT05dsl3aKYqw,23840
-lattifai-1.0.5.dist-info/licenses/LICENSE,sha256=xGMLmdFJy6Jkz3Hd0znyQLmcxC93FSZB5isKnEDMoQQ,1066
-lattifai-1.0.5.dist-info/METADATA,sha256=cTg6ivcixFAv-464qk0R2v19LdEgGkETcNvRzycFSKk,26117
-lattifai-1.0.5.dist-info/WHEEL,sha256=SmOxYU7pzNKBqASvQJ7DjX3XGUF92lrGhMb3R6_iiqI,91
-lattifai-1.0.5.dist-info/entry_points.txt,sha256=F8Akof3VtKtrbnYSav1umgoo9Xbv34rUcKn-ioRfeGQ,474
-lattifai-1.0.5.dist-info/top_level.txt,sha256=tHSoXF26r-IGfbIP_JoYATqbmf14h5NrnNJGH4j5reI,9
-lattifai-1.0.5.dist-info/RECORD,,
+lattifai/workflow/youtube.py,sha256=0B1l_8gdz_O0cy2c9AY9wRPizESQrpRuCP4rwvWRxLA,23687
+lattifai-1.2.0.dist-info/licenses/LICENSE,sha256=xGMLmdFJy6Jkz3Hd0znyQLmcxC93FSZB5isKnEDMoQQ,1066
+lattifai-1.2.0.dist-info/METADATA,sha256=9iEaT3muzKIUmIvQ0oqg4DhM_CvZ53jHvk97kHfPNlQ,37399
+lattifai-1.2.0.dist-info/WHEEL,sha256=SmOxYU7pzNKBqASvQJ7DjX3XGUF92lrGhMb3R6_iiqI,91
+lattifai-1.2.0.dist-info/entry_points.txt,sha256=nHZri2VQkPYEl0tQ0dMYTpVGlCOgVWlDG_JtDR3QXF8,545
+lattifai-1.2.0.dist-info/top_level.txt,sha256=tHSoXF26r-IGfbIP_JoYATqbmf14h5NrnNJGH4j5reI,9
+lattifai-1.2.0.dist-info/RECORD,,

{lattifai-1.0.5.dist-info → lattifai-1.2.0.dist-info}/entry_points.txt

@@ -1,6 +1,7 @@
 [console_scripts]
 lai-align = lattifai.cli.alignment:main
 lai-app-install = lattifai.cli.app_installer:main
+lai-diarize = lattifai.cli.diarization:main
 lai-server = lattifai.cli.server:main
 lai-transcribe = lattifai.cli.transcribe:main
 lai-youtube = lattifai.cli.youtube:main
@@ -11,4 +12,5 @@ laicap-shift = lattifai.cli.caption:main_shift
 [lai_run.cli]
 alignment = lattifai.cli
 caption = lattifai.cli
+diarization = lattifai.cli
 transcribe = lattifai.cli