allenai/Bolmo-7B

Full fine-tune vs allenai/Bolmo-7B-Stage1. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Norms 146/146, Attention 130/130, other 49/49, MLP 111/111, Embeddings 1/2, lm_head 1/1

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse