| { |
| "format": "simpletuner-minimaxmusic-rvq-four-model-comparison-v1", |
| "models": [ |
| { |
| "acoustic_top1": 0.07169886997767858, |
| "acoustic_top5": 0.20938229251217533, |
| "ce_loss": 4.696246407248757, |
| "checkpoint": "checkpoint-17500", |
| "label": "v1, 41M", |
| "loss": 5.337855945933949, |
| "model_id": "SimpleTuner/open-rvq-encoder-minimax-music3-41m-v1", |
| "semantic_top1": 0.41028941761363635, |
| "semantic_top5": 0.7838023792613636, |
| "step": 17500, |
| "teacher_kl_loss": 2.5664437033913354 |
| }, |
| { |
| "acoustic_top1": 0.0762291301094137, |
| "acoustic_top5": 0.21976204453963666, |
| "ce_loss": 4.6332796129877165, |
| "checkpoint": "checkpoint-17500", |
| "label": "v2, 155M", |
| "loss": 5.264569409320809, |
| "model_id": "SimpleTuner/open-rvq-encoder-minimax-music3-155m-v2", |
| "semantic_top1": 0.42864884393063585, |
| "semantic_top5": 0.8017380554552023, |
| "step": 17500, |
| "teacher_kl_loss": 2.525154951679913 |
| }, |
| { |
| "acoustic_top1": 0.07652936662946429, |
| "acoustic_top5": 0.22019304547991073, |
| "ce_loss": 4.629444122314453, |
| "checkpoint": "checkpoint-17500", |
| "label": "v3, 155M + MERT", |
| "loss": 5.260235873135653, |
| "model_id": "SimpleTuner/open-rvq-encoder-minimax-music3-155m-v3", |
| "semantic_top1": 0.4303006258877841, |
| "semantic_top5": 0.8048234419389204, |
| "step": 17500, |
| "teacher_kl_loss": 2.523169430819425 |
| }, |
| { |
| "acoustic_top1": 0.07298060825892858, |
| "acoustic_top5": 0.19882083248782467, |
| "ce_loss": 3.5152001814408735, |
| "checkpoint": "checkpoint-17500", |
| "label": "v4, 169M + depth", |
| "loss": 3.9882268038663, |
| "model_id": "SimpleTuner/open-rvq-encoder-minimax-music3-169m-v4", |
| "semantic_top1": 0.4316433993252841, |
| "semantic_top5": 0.8050786798650568, |
| "step": 17500, |
| "teacher_forced_acoustic_top1": 0.1842243145038555, |
| "teacher_forced_acoustic_top5": 0.44544923460328734, |
| "teacher_forced_semantic_top1": 0.4316433993252841, |
| "teacher_forced_semantic_top5": 0.8050786798650568, |
| "teacher_kl_loss": 1.892104755748402 |
| } |
| ], |
| "notes": [ |
| "Top-k bars use free-running argmax inference for every model.", |
| "V4 teacher-forced loss and acoustic accuracy condition on true earlier codebooks and are not directly comparable to v1-v3 independent-head loss." |
| ], |
| "protocol": { |
| "dataset": "bghira/minimax-music3-rvq-reverse-distillation", |
| "matched_step": 17500, |
| "records": 130, |
| "require_exact_alignment": true, |
| "split": "holdout", |
| "windows": 2768 |
| } |
| } |
|
|
