ALM Data Builder

ALMDataBuilderStage constructs candidate training windows from consecutive diarized audio segments. Each window must meet configurable constraints for duration, sample rate, bandwidth, and speaker count before it is included in the output.

How it Works

For each input manifest entry, the stage:

1. Checks the entry-level sample rate against min_sample_rate (skips the entire entry if below threshold)
2. Iterates over segments as potential window starting points
3. For each starting segment, appends consecutive segments until the accumulated duration reaches the target
4. Optionally truncates the final segment if the window exceeds the maximum duration
5. Validates that the window contains the required number of distinct speakers
6. Records loss statistics for segments that do not meet constraints

The stage processes one AudioTask at a time and mutates the task data in place.

Parameters

Basic Usage

1
from nemo_curator.stages.audio.alm import ALMDataBuilderStage
2
3
# Default configuration: 120-second windows, 2-5 speakers
4
builder = ALMDataBuilderStage()

Advanced Configuration

Short Windows for Fine-Tuning

1
builder = ALMDataBuilderStage(
2
    target_window_duration=30.0,
3
    tolerance=0.2,           # 24-36 seconds
4
    min_speakers=2,
5
    max_speakers=3,
6
)

Permissive Quality Thresholds

1
builder = ALMDataBuilderStage(
2
    min_bandwidth=4000,      # Accept lower-quality audio
3
    min_sample_rate=8000,    # Accept telephone-quality audio
4
    min_speakers=1,          # Allow single-speaker windows
5
    max_speakers=10,
6
)

Preserving Segment Fields

By default, the stage drops words from segments and words,segments from the top level. To preserve all fields:

1
builder = ALMDataBuilderStage(
2
    drop_fields="",
3
    drop_fields_top_level="",
4
)

Output Fields

The stage adds the following fields to each AudioTask:

Speaker Durations

Each window includes a speaker_durations array containing the total speaking time of the top five speakers, sorted by duration in descending order. The array is zero-padded to length five when a window has fewer than five speakers.

Loss Statistics

The stats dictionary contains the following counters:

Best Practices

• Start with the default parameters and adjust based on the stats output
• If lost_spk is high relative to total segments, widen the speaker count range
• If lost_bw is high, verify that the input data has bandwidth metadata in segments[].metrics.bandwidth
• Use drop_fields to reduce output file size when downstream stages do not need word-level or segment-level detail

Related Topics

• ALM Overlap Filtering: Next stage in the ALM pipeline
• ALM Pipeline Concepts: Architectural overview
• ALM Tutorial: End-to-end walkthrough with sample data