RonTon05/Roberta-CLS-URL

Full fine-tune vs papluca/xlm-roberta-base-language-detection. Fine-tune: 100% of the weights changed, across 12 of 12 layers.

Tensors changed: other 50/50, Embeddings 5/5, Norms 48/48, Attention 96/96

1 derivatives checked

• mradermacher/Roberta-CLS-URL-GGUF: Not comparable yet

On the Hub · Downloads on Model Pulse