OpenGVLab/InternVL2_5-8B
Full fine-tune (fewer layers) vs OpenGVLab/InternViT-300M-448px-V2_5. OpenGVLab/InternViT-300M-448px-V2_5 matches part of OpenGVLab/InternVL2_5-8B (340 of its 573 tensors; the rest of OpenGVLab/InternVL2_5-8B has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Embeddings 4/4, Attention 96/96, other 48/48, MLP 96/96, Norms 96/96
5 derivatives checked
- OpenGVLab/InternVL2_5-8B-MPO: Full fine-tune
- OpenGVLab/VisualPRM-8B: Full fine-tune
- ayeshaishaq/DriveLMMo1: Adapter
- OpenGVLab/VisualPRM-8B-v1_1: Full fine-tune
- numind/NuExtract-2-8B-experimental: Full fine-tune
3 other repos hold byte-for-byte identical weight files.