[magpietts] added multiple validation dataloaders and log metrics per val data

[XuesongYang]

Summary

• Cherry picked added multiple validation dataloaders and log metrics per val data. #15189 to the main branch.
• made extra refactors for logging media artifacts.

Details

• Added logging support when running validation on multiple datasets. the training configuration looks like below,

python examples/tts/magpietts.py \\
    ...
    model.train_ds.input_cfg="/data/magpie_pretraining_data/manifests/train_input_cfg_audioCodec21fpsCausalDecoder_en.yaml" \\
    ...
    model.validation_ds.datasets="/xueyang_data/val_input_cfg_audioCodec21fpsCausalDecoder_en.yaml" \\
    ...

where, the yaml config for validation datasets looks like below, which is apt to generalize to multiple languages datasets.

$ cat /xueyang_data/val_input_cfg_audioCodec21fpsCausalDecoder_en.yaml
- name: "LibriTTS_dev_clean"
  input_cfg: "/data/magpie_pretraining_data/manifests/val_input_cfg_audioCodec21fpsCausalDecoder_en.yaml"
- name: "LibriTTS_test_clean"
  input_cfg: "/xueyang_data/val_input_cfg_audioCodec21fpsCausalDecoder_en_LibriTTS_testClean.yaml"

• example plots from wandb:

• wandb log see here: https://wandb.ai/aiapps/debug_magpieTTS_EN_2509/runs/bqerks4y?nw=nwuserxuesong_yang

• The model yaml config that were previously under train_ds.dataset are now directly under 'train_ds'.

• The configuration structure for validation_ds changes:

  • removing the 'dataset' nesting level.
  • requiring a 'datasets' list structure, even for a single validation dataset

[Copilot]

Pull request overview

Adds support for validating MagpieTTS on multiple datasets (multiple validation dataloaders) while improving how media artifacts (audio + attention visualizations) are prepared and logged to W&B/TensorBoard, and updates the example Lhotse config to the new dataset configuration structure.

Changes:

• Refactors validation media logging by separating data preparation (numpy arrays) from logger-specific emission (W&B/TB objects).
• Adds multi-dataloader validation support, including per-dataloader metric aggregation and an averaged validation loss for checkpointing.
• Updates the MagpieTTS Lhotse example config to remove the dataset: nesting and introduce a validation_ds.datasets list format.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.

File	Description
nemo/collections/tts/models/magpietts.py	Implements multi-validation-dataloader handling, refactors media logging, and adjusts Lhotse dataloader config expectations.
examples/tts/conf/magpietts/magpietts_lhotse.yaml	Updates example configuration to match the new train/validation dataset config structure and multi-val datasets list format.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.