Update dataset card for v0.2.0

README.md

@@ -14,360 +14,6 @@ tags:
 - evolutionary-composition
 size_categories:
 - 10K<n<100K
-dataset_info:
-- config_name: base_manifest
-  features:
-  - name: id
-    dtype: string
-  - name: bpm
-    dtype: int64
-  - name: tempo_numerator
-    dtype: float64
-  - name: tempo_denominator
-    dtype: float64
-  - name: key_signature_note
-    dtype: string
-  - name: key_signature_mode
-    dtype: string
-  - name: rainbow_color
-    dtype: string
-  - name: rainbow_color_temporal_mode
-    dtype: string
-  - name: rainbow_color_objectional_mode
-    dtype: string
-  - name: rainbow_color_ontological_mode
-    dtype: string
-  - name: transmigrational_mode
-    dtype: string
-  - name: title
-    dtype: string
-  - name: release_date
-    dtype: string
-  - name: total_running_time
-    dtype: string
-  - name: vocals
-    dtype: bool
-  - name: lyrics
-    dtype: bool
-  - name: lrc_lyrics
-    dtype: string
-  - name: sounds_like
-    dtype: string
-  - name: mood
-    dtype: string
-  - name: genres
-    dtype: string
-  - name: lrc_file
-    dtype: string
-  - name: concept
-    dtype: string
-  - name: training_data
-    struct:
-    - name: album_sequence
-      dtype: int64
-    - name: avg_word_length
-      dtype: float64
-    - name: boundary_fluidity_score
-      dtype: float64
-    - name: concept_length
-      dtype: int64
-    - name: discrepancy_intensity
-      dtype: float64
-    - name: exclamation_marks
-      dtype: int64
-    - name: has_rebracketing_markers
-      dtype: bool
-    - name: memory_discrepancy_severity
-      dtype: float64
-    - name: narrative_complexity
-      dtype: float64
-    - name: ontological_uncertainty
-      dtype: float64
-    - name: question_marks
-      dtype: int64
-    - name: rebracketing_coverage
-      dtype: float64
-    - name: rebracketing_intensity
-      dtype: float64
-    - name: rebracketing_type
-      dtype: string
-    - name: sentence_count
-      dtype: int64
-    - name: temporal_complexity_score
-      dtype: float64
-    - name: track_duration
-      dtype: float64
-    - name: track_id
-      dtype: string
-    - name: track_position
-      dtype: int64
-    - name: uncertainty_level
-      dtype: float64
-    - name: word_count
-      dtype: int64
-  - name: song_structure
-    dtype: string
-  - name: track_id
-    dtype: int64
-  - name: description
-    dtype: string
-  - name: audio_file
-    dtype: string
-  - name: midi_file
-    dtype: string
-  - name: group
-    dtype: string
-  - name: midi_group_file
-    dtype: string
-  - name: player
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 3522729
-    num_examples: 1327
-  download_size: 177786
-  dataset_size: 3522729
-- config_name: training_full
-  features:
-  - name: segment_id
-    dtype: string
-  - name: segment_index
-    dtype: int64
-  - name: song_id
-    dtype: string
-  - name: track_number
-    dtype: int64
-  - name: track_description
-    dtype: string
-  - name: track_group
-    dtype: string
-  - name: track_player
-    dtype: string
-  - name: source_audio_file
-    dtype: string
-  - name: segment_audio_file
-    dtype: string
-  - name: midi_file
-    dtype: string
-  - name: start_seconds
-    dtype: float64
-  - name: end_seconds
-    dtype: float64
-  - name: duration_seconds
-    dtype: float64
-  - name: has_audio
-    dtype: bool
-  - name: has_midi
-    dtype: bool
-  - name: lyric_text
-    dtype: string
-  - name: structure_section
-    dtype: string
-  - name: segment_type
-    dtype: string
-  - name: original_start_seconds
-    dtype: float64
-  - name: original_end_seconds
-    dtype: float64
-  - name: has_structure_adjustments
-    dtype: bool
-  - name: structure_adjustments
-    dtype: string
-  - name: is_sub_segment
-    dtype: bool
-  - name: sub_segment_info
-    dtype: string
-  - name: lrc_line_number
-    dtype: float64
-  - name: lyric_char_count
-    dtype: uint32
-  - name: lyric_word_count
-    dtype: uint32
-  - name: start_adjustment_seconds
-    dtype: float64
-  - name: end_adjustment_seconds
-    dtype: float64
-  - name: content_type
-    dtype: string
-  - name: manifest_track_key
-    dtype: string
-  - name: bpm
-    dtype: int64
-  - name: tempo_numerator
-    dtype: float64
-  - name: tempo_denominator
-    dtype: float64
-  - name: key_signature_note
-    dtype: string
-  - name: key_signature_mode
-    dtype: string
-  - name: rainbow_color
-    dtype: string
-  - name: rainbow_color_temporal_mode
-    dtype: string
-  - name: rainbow_color_objectional_mode
-    dtype: string
-  - name: rainbow_color_ontological_mode
-    dtype: string
-  - name: transmigrational_mode
-    dtype: string
-  - name: title
-    dtype: string
-  - name: release_date
-    dtype: string
-  - name: total_running_time
-    dtype: string
-  - name: vocals
-    dtype: bool
-  - name: lyrics
-    dtype: bool
-  - name: lrc_lyrics
-    dtype: string
-  - name: sounds_like
-    dtype: string
-  - name: mood
-    dtype: string
-  - name: genres
-    dtype: string
-  - name: lrc_file
-    dtype: string
-  - name: concept
-    dtype: string
-  - name: training_data
-    struct:
-    - name: album_sequence
-      dtype: int64
-    - name: avg_word_length
-      dtype: float64
-    - name: boundary_fluidity_score
-      dtype: float64
-    - name: concept_length
-      dtype: int64
-    - name: discrepancy_intensity
-      dtype: float64
-    - name: exclamation_marks
-      dtype: int64
-    - name: has_rebracketing_markers
-      dtype: bool
-    - name: memory_discrepancy_severity
-      dtype: float64
-    - name: narrative_complexity
-      dtype: float64
-    - name: ontological_uncertainty
-      dtype: float64
-    - name: question_marks
-      dtype: int64
-    - name: rebracketing_coverage
-      dtype: float64
-    - name: rebracketing_intensity
-      dtype: float64
-    - name: rebracketing_type
-      dtype: string
-    - name: sentence_count
-      dtype: int64
-    - name: temporal_complexity_score
-      dtype: float64
-    - name: track_duration
-      dtype: float64
-    - name: track_id
-      dtype: string
-    - name: track_position
-      dtype: int64
-    - name: uncertainty_level
-      dtype: float64
-    - name: word_count
-      dtype: int64
-  - name: song_structure
-    dtype: string
-  - name: midi_group_file
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 35869494
-    num_examples: 11605
-  download_size: 580351
-  dataset_size: 35869494
-- config_name: training_segments
-  features:
-  - name: segment_id
-    dtype: string
-  - name: segment_index
-    dtype: int64
-  - name: track_id
-    dtype: string
-  - name: track_number
-    dtype: int64
-  - name: track_description
-    dtype: string
-  - name: track_group
-    dtype: string
-  - name: track_player
-    dtype: string
-  - name: source_audio_file
-    dtype: string
-  - name: segment_audio_file
-    dtype: string
-  - name: midi_file
-    dtype: string
-  - name: start_seconds
-    dtype: float64
-  - name: end_seconds
-    dtype: float64
-  - name: duration_seconds
-    dtype: float64
-  - name: has_audio
-    dtype: bool
-  - name: has_midi
-    dtype: bool
-  - name: lyric_text
-    dtype: string
-  - name: structure_section
-    dtype: string
-  - name: segment_type
-    dtype: string
-  - name: original_start_seconds
-    dtype: float64
-  - name: original_end_seconds
-    dtype: float64
-  - name: has_structure_adjustments
-    dtype: bool
-  - name: structure_adjustments
-    dtype: string
-  - name: is_sub_segment
-    dtype: bool
-  - name: sub_segment_info
-    dtype: string
-  - name: lrc_line_number
-    dtype: float64
-  - name: lyric_char_count
-    dtype: uint32
-  - name: lyric_word_count
-    dtype: uint32
-  - name: start_adjustment_seconds
-    dtype: float64
-  - name: end_adjustment_seconds
-    dtype: float64
-  - name: content_type
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 6049737
-    num_examples: 11605
-  download_size: 389389
-  dataset_size: 6049737
-configs:
-- config_name: base_manifest
-  data_files:
-  - split: train
-    path: base_manifest/train-*
-- config_name: training_full
-  data_files:
-  - split: train
-    path: training_full/train-*
-- config_name: training_segments
-  data_files:
-  - split: train
-    path: training_segments/train-*
 ---
 
 # White Training Data
@@ -388,7 +34,7 @@ These models are **fitness functions for evolutionary music composition**, not c
 
 ## Version
 
-Current: **v0.2.0** — 2026-02-10
+Current: **v0.2.0** — 2026-02-12
 
 ## Dataset Structure
 
@@ -407,11 +53,11 @@ Current: **v0.2.0** — 2026-02-10
 | Orange | 1,731 | 83.8% | 51.1% | 100.0% |
 | Yellow | 656 | 88.0% | 52.9% | 52.6% |
 | Green | 393 | 90.1% | 69.5% | 0.0% |
-| Blue | 2,097 | 96.0% | 12.1% | 100.0% |
-| Indigo | 1,406 | 77.2% | 34.1% | 100.0% |
 | Violet | 2,100 | 75.9% | 55.6% | 100.0% |
+| Indigo | 1,406 | 77.2% | 34.1% | 100.0% |
+| Blue | 2,097 | 96.0% | 12.1% | 100.0% |
 
-**Note:** Audio waveforms and MIDI binaries are stored separately (not included in this dataset due to size). This dataset contains the metadata, segment boundaries, lyric text, and computed training features needed for model training. The media parquet (~15 GB) is used locally during training.
+**Note:** Audio waveforms and MIDI binaries are stored separately (not included in metadata configs due to size). The `preview` config includes playable audio for exploration. The media parquet (~15 GB) is used locally during training.
 
 ## Key Features
 
@@ -426,6 +72,12 @@ Current: **v0.2.0** — 2026-02-10
 - `has_audio` / `has_midi` — Modality availability flags
 - `start_seconds` / `end_seconds` — Segment time boundaries
 
+### `preview` (playable audio)
+
+Same metadata fields as `training_full`, plus:
+- `audio` — Audio feature with inline playback support (FLAC encoded, 44.1kHz)
+- `duration_seconds` — Segment duration
+
 ## Usage
 
 ```python
@@ -434,6 +86,9 @@ from datasets import load_dataset
 # Load the primary training table (segments + manifest metadata)
 training = load_dataset("earthlyframes/white-training-data", "training_full")
 
+# Load playable audio preview
+preview = load_dataset("earthlyframes/white-training-data", "preview")
+
 # Load just the base manifest (track-level)
 manifest = load_dataset("earthlyframes/white-training-data", "base_manifest")
 
@@ -465,4 +120,4 @@ Spatial mode is bottlenecked by instrumental albums (Yellow, Green) which lack t
 [Collaborative Intelligence License v1.0](https://github.com/brotherclone/white/blob/main/COLLABORATIVE_INTELLIGENCE_LICENSE.md) — This work represents conscious partnership between human creativity and AI. Both parties have agency; both must consent to sharing.
 
 ---
-*Generated 2026-02-10 | [GitHub](https://github.com/brotherclone/white)*
+*Generated 2026-02-12 | [GitHub](https://github.com/brotherclone/white)*