devon7y/Toronto-Mans-35B-A3B-FP8
Full fine-tune vs devon7y/Toronto-Mans-35B-A3B. Fine-tune: 93% of the weights changed, across 40 of 40 layers.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/131, Attention 0/280, MLP 80/280
Full fine-tune vs devon7y/Toronto-Mans-35B-A3B. Fine-tune: 93% of the weights changed, across 40 of 40 layers.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/131, Attention 0/280, MLP 80/280