Hugging Face's logo

arda-argmax
/
sortformer-finetuned

NeMo
Model card Files
xet
 Community
sortformer-finetuned / 6spk-fullrandom-bs10-ga8-a100-80gb-1gpu-lr5e-4-lr-v1-1781045292 /README.md
arda-argmax's picture
arda-argmax
Init run 6spk-fullrandom-bs10-ga8-a100-80gb-1gpu-lr5e-4-lr-v1-1781045292
5267104 verified
preview code
|
Raw
History Blame Contribute Delete
878 Bytes
metadata 
library_name: nemo
tags:
  - sortformer
  - speaker-diarization
  - automatic-speech-recognition

6spk-fullrandom-bs10-ga8-a100-80gb-1gpu-lr5e-4-lr-v1-1781045292

Sortformer fine-tuning run, hosted under arda-argmax/sortformer-finetuned.

Best checkpoint

  • monitor: val_der (min)

Files:

  • best.nemo — full NeMo checkpoint (config + weights), restorable via SortformerEncLabelModel.restore_from('best.nemo')
  • best.json — machine-readable metadata for the best checkpoint
  • checkpoints/step_*_epoch_*_*.nemo — archived copies of every checkpoint that became the best during training
  • config.yaml — Hydra config the run was launched with

Run info