### Install ForceAlign using pip

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Installs the ForceAlign Python library using pip. Ensure pip3 is available and ffmpeg is installed separately for audio processing.

```bash
pip3 install forcealign
```

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

### ForceAlign Class Initialization and Inference (Python)

Source: https://context7.com/lukerbs/forcealign/llms.txt

Demonstrates initializing the ForceAlign class with an audio file and transcript, running inference to get word and phoneme alignments, and accessing the raw text. Supports .mp3 and .wav audio formats.

```python
from forcealign import ForceAlign

# Initialize with audio file and transcript
transcript = "The quick brown fox jumps over the lazy dog."
align = ForceAlign(audio_file='./speech.mp3', transcript=transcript)

# Run alignment inference - returns list of Word objects
words = align.inference()

# Access word-level alignments
for word in words:
    print(f"Word: {word.word}, Start: {word.time_start}s, End: {word.time_end}s")
    # Output: Word: THE, Start: 0.241s, End: 0.361s
    #         Word: QUICK, Start: 0.381s, End: 0.602s
    #         ...

# Access phoneme-level alignments for each word
for word in words:
    print(f"\nWord: {word.word}")
    for phoneme in word.phonemes:
        print(f"  Phoneme: {phoneme}, Start: {word.time_start}s")
    # Output: Word: THE
    #           Phoneme: DH, Start: 0.241s
    #           Phoneme: AH0, Start: 0.241s

# Access the raw transcript text
print(align.raw_text)  # "The quick brown fox jumps over the lazy dog."

# Access all phoneme alignments
for phoneme in align.phoneme_alignments:
    print(f"Phoneme: {phoneme.phoneme}, Start: {phoneme.time_start}s, End: {phoneme.time_end}s")
```

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

### Accessing Phoneme Alignments in Python

Source: https://context7.com/lukerbs/forcealign/llms.txt

Shows how to iterate through phoneme alignments obtained from the ForceAlign inference. Each phoneme object includes its phonetic representation and precise start and end times, crucial for detailed speech analysis.

```python
from forcealign import ForceAlign

align = ForceAlign(audio_file='./speech.mp3', transcript="Hello")
words = align.inference()

# Access phoneme alignments
for phoneme in align.phoneme_alignments:
    print(f"Phoneme: {phoneme.phoneme}")
    print(f"Start: {phoneme.time_start}s")
    print(f"End: {phoneme.time_end}s")
    # Output: Phoneme: HH
    #         Start: 0.241s
    #         End: 0.305s
    #         Phoneme: AH0
    #         Start: 0.305s
    #         End: 0.369s
    #         ...

# String representation
print(repr(align.phoneme_alignments[0]))
```

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

### Get Word-Level Alignments with Transcript (Python)

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Performs word-level forced alignment using a provided audio file and transcript. It initializes ForceAlign with the audio path and transcript, then runs inference to get word timings.

```python
from forcealign import ForceAlign

# Provide path to audio file and corresponding transcript
transcript = "The quick brown fox jumps over the lazy dog."
align = ForceAlign(audio_file='./speech.mp3', transcript=transcript)

# Run prediction and return alignment results
words = align.inference()

# Show predicted word-level alignments
for word in words:
    print(f"Word: {word.word}, Start: {word.time_start}s, End: {word.time_end}s")
```

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

### Accessing Word Attributes in Python

Source: https://context7.com/lukerbs/forcealign/llms.txt

Demonstrates how to access attributes of a 'word' object, such as its text, start and end times, phonemes, and breath detection status. This is useful for analyzing individual words in an aligned transcript.

```python
word = words[0]
print(f"Word text: {word.word}")           # "HELLO"
print(f"Start time: {word.time_start}s")   # 0.241
print(f"End time: {word.time_end}s")       # 0.562
print(f"Phonemes: {word.phonemes}")        # ['HH', 'AH0', 'L', 'OW1']
print(f"Has breath after: {word.breath}")  # True/False (detected from punctuation)

# String representation
print(repr(word))
```

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

### ForceAlign with Automatic Transcription (Python)

Source: https://context7.com/lukerbs/forcealign/llms.txt

Shows how to use ForceAlign without providing a transcript. The library automatically generates a transcript using Wav2Vec2 speech-to-text before performing alignment. Supports .mp3 and .wav audio formats.

```python
from forcealign import ForceAlign

# Initialize without transcript - automatic speech-to-text
align = ForceAlign(audio_file='./speech.mp3')
# Output: No transcript provided. Generating transcript using speech_to_text...
#         Generated Transcript: THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG

# Run alignment inference
words = align.inference()

# Access the auto-generated transcript
print(f"Generated Transcript: {align.raw_text}")
# Output: Generated Transcript: THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG

# Word alignments work the same way
for word in words:
    print(f"Word: {word.word}, Start: {word.time_start}s, End: {word.time_end}s")
```

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

### Reviewing Alignment with Audio Playback (Python)

Source: https://context7.com/lukerbs/forcealign/llms.txt

Utilizes the `review_alignment` method to play an audio file while printing word alignments in real-time to the console. This aids in visually verifying the accuracy of the forced alignment. Requires inference to be run first.

```python
from forcealign import ForceAlign

transcript = "The dog quickly jumped over the large cat."
align = ForceAlign(audio_file='./speech.mp3', transcript=transcript)

# Must run inference first to generate alignments
words = align.inference()

# Play audio and print words as they are spoken
align.review_alignment()
# Output (printed in real-time as audio plays):
#   THE: 0.241 -- 0.361(s)
#   DOG: 0.381 -- 0.522(s)
#   QUICKLY: 0.542 -- 0.843(s)
#   ...

```

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

### Review Word-Level Alignments in Real-Time (Python)

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Enables real-time review of word-level alignments by playing the audio file concurrently with printing the alignment results. This is useful for immediate verification and debugging.

```python
from forcealign import ForceAlign

# Provide path to audio file and transcript
transcript = "The quick brown fox jumps over the lazy dog."
align = ForceAlign(audio_file='./speech.mp3', transcript=transcript)

# Play the audio while printing word alignments in real-time
align.review_alignment()
```

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

### Standalone Speech-to-Text Transcription (Python)

Source: https://context7.com/lukerbs/forcealign/llms.txt

Demonstrates the `speech_to_text` function for direct audio-to-text conversion using the Wav2Vec2 model, independent of the alignment process. It supports both .mp3 and .wav files and handles FileNotFoundError for missing audio files.

```python
from forcealign import speech_to_text

# Convert audio to text
transcript = speech_to_text("./speech.mp3")
print(transcript)
# Output: THE QUICK BROWN FOX JUMPED OVER THE LAZY DOG

# Works with both .mp3 and .wav files
transcript_wav = speech_to_text("./speech.wav")
print(transcript_wav)
# Output: THE QUICK BROWN FOX JUMPED OVER THE LAZY DOG

# Raises FileNotFoundError if audio file doesn't exist
try:
    transcript = speech_to_text("./nonexistent.mp3")
except FileNotFoundError as e:
    print(f"Error: {e}")
    # Output: Error: Audio file not found: ./nonexistent.mp3

```

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

### Speech-to-Text Conversion (Python)

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Utilizes ForceAlign as a standalone speech-to-text tool to generate a transcript directly from an audio file. This function is useful when only transcription is needed.

```python
from forcealign import speech_to_text

# Generate transcript directly from an audio file
generated_transcript = speech_to_text("./speech.mp3")
print(generated_transcript)
# Output: "THE QUICK BROWN FOX JUMPED OVER THE LAZY DOG"
```

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

### Word-Level Alignments with Auto Transcription (Python)

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Performs word-level forced alignment by automatically generating a transcript from the audio file using Wav2Vec2. It initializes ForceAlign with only the audio path and then runs inference.

```python
from forcealign import ForceAlign

# Provide path to audio file; omit transcript
align = ForceAlign(audio_file='./speech.mp3')

# Automatically generate transcript and align words
words = align.inference()

# Show the generated transcript
print("Generated Transcript:")
print(align.raw_text)

# Show predicted word-level alignments
for word in words:
    print(f"Word: {word.word}, Start: {word.time_start}s, End: {word.time_end}s")
```

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

### Phoneme-Level Alignments (Python)

Source: https://github.com/lukerbs/forcealign/blob/main/README.md

Performs phoneme-level forced alignment using a provided audio file and transcript. It retrieves word alignments and then iterates through the phonemes within each word to display their timings.

```python
from forcealign import ForceAlign

# Provide path to audio file and transcript
transcript = "The quick brown fox jumps over the lazy dog."
align = ForceAlign(audio_file='./speech.mp3', transcript=transcript)

# Run prediction and return alignment results
words = align.inference()

# Access predicted phoneme-level alignments
for word in words:
    print(f"Word: {word.word}")
    for phoneme in word.phonemes:
        print(f"Phoneme: {phoneme.phoneme}, Start: {phoneme.time_start}s, End: {phoneme.time_end}s")
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.