YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

moe_shared

Decoder-only transformer for Vietnamese/Japanese to English translation. Feed-forward variant: moe_shared.

  • Dataset: local:data/en-vi-ja-curated-500k-triplets
  • Total parameters: 29,280,232
  • Active parameters: 22,988,788
  • Expert hidden size: 511
  • Train tokens: 30,000,000

The checkpoint is best.pt (lowest validation cross-entropy).

Downloads last month
9
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support