lmms-lab/EgoGPT-7b-EgoIT
Partial fine-tune (fewer layers) vs lmms-lab/llava-onevision-qwen2-7b-ov. lmms-lab/llava-onevision-qwen2-7b-ov matches part of lmms-lab/EgoGPT-7b-EgoIT (765 of its 1256 tensors; the rest of lmms-lab/EgoGPT-7b-EgoIT has no counterpart). On the shared part: Partial change: 88% of the weights changed (attention, lm_head, mlp, norms, other), in 28 of 28 layers.
Tensors changed: Embeddings 0/4, other 5/5, Norms 54/163, MLP 84/188, Attention 196/404, lm_head 1/1
1 other repos hold byte-for-byte identical weight files.