SimpleTuner/open-rvq-encoder-minimax-music3-41m-v1
Updated • 1
Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
schema_version: int64
model: string
model_version: string
return_full_prediction: bool
gpu_replicas: int64
jobs: list<item: struct<status: string, id: string, index: int64, prompt: string, lyrics: string, seed: in (... 678 chars omitted)
child 0, item: struct<status: string, id: string, index: int64, prompt: string, lyrics: string, seed: int64, reques (... 666 chars omitted)
child 0, status: string
child 1, id: string
child 2, index: int64
child 3, prompt: string
child 4, lyrics: string
child 5, seed: int64
child 6, requested_duration_seconds: double
child 7, num_inference_steps: int64
child 8, sampling_rate: int64
child 9, frame_rate: int64
child 10, latent_hop_length: int64
child 11, source_metadata: struct<title: string, source: string, source_detail: string, source_category: string, source_url: st (... 167 chars omitted)
child 0, title: string
child 1, source: string
child 2, source_detail: string
child 3, source_category: string
child 4, source_url: string
child 5, handle: string
child 6, created_at: string
child 7, model_name: string
child 8, major_model_version: string
child 9, duration: double
child 10, play_count: int64
child 11, upvote_count: int64
child 12, audio_url: string
child 12, audio_file: string
child 13, priming_frames: int64
child 14, em
...
double>
child 0, temperature: double
child 1, flow_guidance_scale: double
child 1, low: struct<temperature: double, flow_guidance_scale: double>
child 0, temperature: double
child 1, flow_guidance_scale: double
child 2, high: struct<temperature: double, flow_guidance_scale: double>
child 0, temperature: double
child 1, flow_guidance_scale: double
holdout_fraction: double
audio_duration: double
sampling_variation_fraction: double
alignment_convention: struct<priming_code_row: int64, emitted_code_row_offset: int64, nominal_flow_latents_per_semantic_fr (... 45 chars omitted)
child 0, priming_code_row: int64
child 1, emitted_code_row_offset: int64
child 2, nominal_flow_latents_per_semantic_frame: double
child 3, canonical_exact_mapping: string
seed: int64
selected_jobs: int64
instrumental_jobs: int64
legacy_default_sampling: struct<autoregressive_temperature: double, autoregressive_sampling_top_k: int64, autoregressive_guid (... 173 chars omitted)
child 0, autoregressive_temperature: double
child 1, autoregressive_sampling_top_k: int64
child 2, autoregressive_guidance_scale: double
child 3, autoregressive_guidance_top_k: int64
child 4, flow_guidance_scale: double
child 5, flow_num_inference_steps: int64
child 6, flow_scheduler: string
child 7, flow_scheduler_shift: double
skipped: struct<missing: int64, duplicate: int64, too_long: int64>
child 0, missing: int64
child 1, duplicate: int64
child 2, too_long: int64
to
{'schema_version': Value('int64'), 'model': Value('string'), 'jobs_per_shard': Value('int64'), 'audio_duration': Value('float64'), 'num_inference_steps': Value('int64'), 'seed': Value('int64'), 'selected_jobs': Value('int64'), 'skipped': {'missing': Value('int64'), 'duplicate': Value('int64'), 'too_long': Value('int64')}, 'holdout_fraction': Value('float64'), 'holdout_hash_salt': Value('string'), 'sampling_variation_fraction': Value('float64'), 'sampling_variants': {'default': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}, 'low': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}, 'high': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}}, 'legacy_default_sampling': {'autoregressive_temperature': Value('float64'), 'autoregressive_sampling_top_k': Value('int64'), 'autoregressive_guidance_scale': Value('float64'), 'autoregressive_guidance_top_k': Value('int64'), 'flow_guidance_scale': Value('float64'), 'flow_num_inference_steps': Value('int64'), 'flow_scheduler': Value('string'), 'flow_scheduler_shift': Value('float64')}, 'lyric_bearing_jobs': Value('int64'), 'instrumental_jobs': Value('int64'), 'hub_repo_id': Value('string'), 'alignment_convention': {'priming_code_row': Value('int64'), 'emitted_code_row_offset': Value('int64'), 'nominal_flow_latents_per_semantic_frame': Value('float64'), 'canonical_exact_mapping': Value('string')}}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
schema_version: int64
model: string
model_version: string
return_full_prediction: bool
gpu_replicas: int64
jobs: list<item: struct<status: string, id: string, index: int64, prompt: string, lyrics: string, seed: in (... 678 chars omitted)
child 0, item: struct<status: string, id: string, index: int64, prompt: string, lyrics: string, seed: int64, reques (... 666 chars omitted)
child 0, status: string
child 1, id: string
child 2, index: int64
child 3, prompt: string
child 4, lyrics: string
child 5, seed: int64
child 6, requested_duration_seconds: double
child 7, num_inference_steps: int64
child 8, sampling_rate: int64
child 9, frame_rate: int64
child 10, latent_hop_length: int64
child 11, source_metadata: struct<title: string, source: string, source_detail: string, source_category: string, source_url: st (... 167 chars omitted)
child 0, title: string
child 1, source: string
child 2, source_detail: string
child 3, source_category: string
child 4, source_url: string
child 5, handle: string
child 6, created_at: string
child 7, model_name: string
child 8, major_model_version: string
child 9, duration: double
child 10, play_count: int64
child 11, upvote_count: int64
child 12, audio_url: string
child 12, audio_file: string
child 13, priming_frames: int64
child 14, em
...
double>
child 0, temperature: double
child 1, flow_guidance_scale: double
child 1, low: struct<temperature: double, flow_guidance_scale: double>
child 0, temperature: double
child 1, flow_guidance_scale: double
child 2, high: struct<temperature: double, flow_guidance_scale: double>
child 0, temperature: double
child 1, flow_guidance_scale: double
holdout_fraction: double
audio_duration: double
sampling_variation_fraction: double
alignment_convention: struct<priming_code_row: int64, emitted_code_row_offset: int64, nominal_flow_latents_per_semantic_fr (... 45 chars omitted)
child 0, priming_code_row: int64
child 1, emitted_code_row_offset: int64
child 2, nominal_flow_latents_per_semantic_frame: double
child 3, canonical_exact_mapping: string
seed: int64
selected_jobs: int64
instrumental_jobs: int64
legacy_default_sampling: struct<autoregressive_temperature: double, autoregressive_sampling_top_k: int64, autoregressive_guid (... 173 chars omitted)
child 0, autoregressive_temperature: double
child 1, autoregressive_sampling_top_k: int64
child 2, autoregressive_guidance_scale: double
child 3, autoregressive_guidance_top_k: int64
child 4, flow_guidance_scale: double
child 5, flow_num_inference_steps: int64
child 6, flow_scheduler: string
child 7, flow_scheduler_shift: double
skipped: struct<missing: int64, duplicate: int64, too_long: int64>
child 0, missing: int64
child 1, duplicate: int64
child 2, too_long: int64
to
{'schema_version': Value('int64'), 'model': Value('string'), 'jobs_per_shard': Value('int64'), 'audio_duration': Value('float64'), 'num_inference_steps': Value('int64'), 'seed': Value('int64'), 'selected_jobs': Value('int64'), 'skipped': {'missing': Value('int64'), 'duplicate': Value('int64'), 'too_long': Value('int64')}, 'holdout_fraction': Value('float64'), 'holdout_hash_salt': Value('string'), 'sampling_variation_fraction': Value('float64'), 'sampling_variants': {'default': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}, 'low': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}, 'high': {'temperature': Value('float64'), 'flow_guidance_scale': Value('float64')}}, 'legacy_default_sampling': {'autoregressive_temperature': Value('float64'), 'autoregressive_sampling_top_k': Value('int64'), 'autoregressive_guidance_scale': Value('float64'), 'autoregressive_guidance_top_k': Value('int64'), 'flow_guidance_scale': Value('float64'), 'flow_num_inference_steps': Value('int64'), 'flow_scheduler': Value('string'), 'flow_scheduler_shift': Value('float64')}, 'lyric_bearing_jobs': Value('int64'), 'instrumental_jobs': Value('int64'), 'hub_repo_id': Value('string'), 'alignment_convention': {'priming_code_row': Value('int64'), 'emitted_code_row_offset': Value('int64'), 'nominal_flow_latents_per_semantic_frame': Value('float64'), 'canonical_exact_mapping': Value('string')}}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
MiniMax Music 3 RVQ Reverse-Distillation Traces
Research traces generated from MiniMax Music 3. Each ZIP contains generated
audio, sampled RVQ codes, teacher sampling logits, conditioning embeddings,
Flow-VAE latents, and source prompt metadata. See campaign-config.json for
generation settings and indexes/ for per-commit manifests.
Dataset statistics
As of 2026-08-16, the uploaded snapshot contains:
- 2,972 generated songs
- 91.84 hours of actual generated audio
- 1 minute 51 seconds mean duration
- 1 minute 59 seconds median duration
- 2,365 songs (79.6%) that ended naturally with the semantic EOS code
Generation requests were capped at 180 seconds, representing 148.6 requested
hours before early stopping. Actual durations are calculated from
flow_latent_length * latent_hop_length / sampling_rate, using the values stored
for every uploaded sample.
The canonical metadata lives in indexes/. It includes exact or legacy-default
sampling parameters and deterministic train/holdout assignments. New shards
sample a deterministic mix of temperature and flow-CFG settings; legacy shards
are explicitly marked with the original fixed defaults.
RVQ row 0 is an un-emitted priming row; emitted semantic frame i is
codes[i + 1]. The nominal Flow-VAE/semantic ratio is 3.4453125, but consumers
must use the stored chunk alignment where available because chunk-boundary
rounding is not uniform. Legacy indexes identify records where exact chunk
spans were not captured.
For encoder training, re-encode audio.flac with the full upstream Flow-VAE.
The captured flow_vae_latents are decoder-path consistency targets, not a
replacement for audio-domain training inputs.
- Downloads last month
- 3,157
