OpenGVLab/InternVL3_5-8B-Flash
Full fine-tune vs OpenGVLab/InternVL3_5-8B. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: MLP 210/210, Norms 231/241, Attention 240/240, Embeddings 5/5, other 48/48, lm_head 1/1
Full fine-tune vs OpenGVLab/InternVL3_5-8B. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: MLP 210/210, Norms 231/241, Attention 240/240, Embeddings 5/5, other 48/48, lm_head 1/1