fpadovani/eng-100mb-after-wc-uniform-newlex-eng-ckpt500_seed10_seed10

Full fine-tune vs fpadovani/ppt-wc-uniform-newlex-eng-100mb_seed10. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: Attention 48/48, Norms 25/50, MLP 48/48, Embeddings 1/1

On the Hub · Downloads on Model Pulse